MLinfo | 機械学習・AI論文まとめ

Learning Length-Extrapolatable Recurrent Models

Recurrent models provide a natural path to long-context modeling, yet models trained with backpropagation thro

深層学習RNN / LSTMテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

When Does Scale-Invariant Optimization Become Unstable? An Exact Schedule Law with Weight Decay

Normalization renders large parts of neural networks effectively scale invariant, inducing a hidden feedback l

深層学習CNN回帰テキスト

用途: 回帰
難易度: Hard
コスト: High

説明可能品質予測/異常検知深層学習Transformer生成テキスト

SAEScientist-Bench: Can AI Agents Conduct Autonomous SAE Interpretability Research?

While research on recursive self-improvement (RSI) has predominantly automated model training pipelines, relia

用途: 生成
難易度: Hard
コスト: High

Curriculum Learning as Transport: Understanding Curricula with Wasserstein Geodesics

Curriculum learning is governed by several coupled design choices---how difficulty is defined, how examples ar

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

PlayTrain: An Efficient Reinforcement Learning Framework for LLM-Generated Adaptable JavaScript Games

While many video-game environments (VGEs) have played crucial roles in advancing reinforcement learning (RL),

深層学習軽量化・量子化生成画像テキスト

用途: 生成
難易度: Hard
コスト: High

Training-Free Task Vectors for LLM Behavioral Control

Task vectors enable post-training model editing by identifying semantically meaningful directions in weight sp

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Do Reasoning Representations Help Humans Evaluate LLM Outputs?

Reasoning representations are increasingly used as explanations for large language model outputs. Yet they are

説明可能自然言語処理大規模言語モデル検出テキスト

用途: 検出
難易度: Hard
コスト: High

Answer-Distribution Trajectories: A Stochastic-Dynamics View of LLM Reasoning

Chain-of-thought reasoning provides a structured computation between a model's input and final answer. Yet it

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Evaluation of Contextual Understanding in Large Language Models

Large Language Models (LLMs) demonstrate impressive performance across diverse NLP tasks, yet their ability to

説明可能自然言語処理大規模言語モデルQAテキスト

用途: QA
難易度: Hard
コスト: High

Transformers as In-Context Samplers: From Closed-Form Diffusion to Estimation-Free Sampling

A growing body of work establishes that large language models are not mere statistical memorizers, but are cap

深層学習Transformer生成回帰テキスト

用途: 生成
難易度: Hard
コスト: High

Omni Interaction Agent Technical Report

In this work, we present Gander, an end-to-end model that unifies omni perception, realtime interaction, and a

生成AIGANテキスト音声動画

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

コンピュータビジョンセグメンテーション画像テキスト

The BatchNorm Illusion: Diagnosing Normalization Artifacts in Machine Unlearning Evaluation

Approximate machine unlearning aims to remove the influence of specific training data from a trained model wit

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Chimaera: A Mixture-of-Graph-Experts Architecture for Cross-Task and Cross-Dataset Graph Learning

Designing foundation models for graphs is challenging due to the irregular structure of graphs and the differe

深層学習グラフニューラルネット分類テキスト

用途: 分類
難易度: Hard
コスト: High

センサ/時系列品質予測/異常検知深層学習Transformer生成テキスト音声

TontaubeV1: Streaming Text-to-Speech with Hierarchical Codec Modeling and Bounded Context

Text-to-speech systems often face a trade-off between natural prosody and efficient inference: higher perceptu

用途: 生成
難易度: Hard
コスト: High

説明可能品質予測/異常検知自然言語処理大規模言語モデルテキスト

Suan: Rectifying Direct Preference Safety Alignment in Large Language Models

Integrating robust safety guardrails into Large Language Models (LLMs) is essential for delivering helpful yet

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Leveraging contextual events on structure-aware next activity prediction

Predictive process monitoring aims at forecasting various aspects of running processes. Among the different ta

深層学習Transformer分類予測テキスト

用途: 分類
難易度: Hard
コスト: Low

Certified Topological Interaction in Neural Representations: Class Disentanglement Is Mostly Pairwise

Class disentanglement (the separation of a representation's class-conditional point clouds along depth and ove

深層学習正規化・最適化手法画像テキスト3D

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

表形式向き深層学習軽量化・量子化画像テキスト表形式

When Topology Betrays Privacy: Lattice-Based Reconstruction Attacks on Secure Aggregation in Decentralized Federated Learning

Secure Aggregation (SA) is widely regarded as a strong defense against model-update leakage in Federated Learn

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Do Reviewers Still Reward Lexical Complexity? A Frozen-Rater Study of Preference Drift in 124K ICLR Reviews

Large language models have collapsed the cost of producing lexically elaborate prose, and whether peer reviewe

MI向き深層学習軽量化・量子化回帰テキスト

用途: 回帰
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデルテキスト強化学習

Environments as Scaffold: Enriching Feedback to Bootstrap Self-Evolving Agents in Long-Horizon Tasks

Large Language Models demonstrate remarkable proficiency in static reasoning, yet training them as autonomous

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

コンピュータビジョン物体検出テキストマルチモーダル

To Adapt or Not to Adapt? Selective Adaptation for Vision-Language Models

Test-time adaptation (TTA) has emerged as a prominent strategy for adapting vision-language models to distribu

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

TV-Regulated OPD: Direction Matters in On-Policy Distillation

On-Policy Distillation (OPD) facilitates the transfer of knowledge from domain expert to student in the post-t

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

EMBLEM: Enhancing Multi-script Table Detection through Masking

Table detection is a core task in document analysis, supporting downstream applications such as information re

自然言語処理ファインチューニング検出QA検索

用途: 検出
難易度: Hard
コスト: High

CPUで試しやすい品質予測/異常検知深層学習軽量化・量子化テキスト

A Measurement Study of LLM Inference Trade-offs Across Edge Continuum Hardware

Large language models (LLMs) are increasingly used as backends for intelligent web services, but serving them

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

HoneyRoute: Honeypot-Model Routing for Adversarial LLM Serving

We introduce HoneyRoute, an inference-serving layer that detects whether an incoming request is malicious and,

自然言語処理大規模言語モデル検出テキスト

用途: 検出
難易度: Hard
コスト: High

品質予測/異常検知深層学習RNN / LSTMテキスト強化学習

Routing Dense Layouts with History-Aware Offline Reinforcement Learning using LSTM

Detailed routing remains a dominant runtime bottleneck in physical design due to increasing complexity of desi

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

Vision: Data-Centric Anchoring for Robust and Interpretable Agentic AI

Agentic AI systems built on large language models fail in two persistent ways that scaling does not fix: they

説明可能自然言語処理大規模言語モデル生成テキスト

用途: 生成
難易度: Hard
コスト: High

説明可能品質予測/異常検知深層学習Transformer検出生成異常検知

SIM: Subspace Interaction-based Method for Token-Level Text Anomaly Detection

Token-level text anomaly detection, as an emerging trend of text anomaly detection, moves beyond coarse-graine

用途: 検出
難易度: Easy
コスト: Low

コンピュータビジョンセグメンテーション生成画像テキスト

Geodesic-informed Generative Diffusion Model For Topology-preserved Image Video Generation

Generative diffusion models have emerged as a class of powerful techniques for various imaging applications, i

用途: 生成
難易度: Hard
コスト: High

KBBQ: A Predictive Noise Law and the Limits of Spectrum Flattening in FP4 Quantization

We develop a second-order theory of quantization noise in matrix multiplication in which the quantization form

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

DISEIL: Demonstration Distillation for Sample-Efficient Imitation Learning

A robot that can be taught a new task from a handful of demonstrations has to work out for itself what it stil

深層学習軽量化・量子化画像テキストマルチモーダル

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

MI向き品質予測/異常検知コンピュータビジョンセグメンテーション分類生成テキスト

PocketVE: Stable and Property-Guided Structure-Based Drug Design with Variance-Exploding Diffusion

Protein-conditioned 3D molecule generation is a central challenge in structure-based drug design, requiring a

用途: 分類
難易度: Hard
コスト: High

Proactive Context-Forecasted Safety Constraints for Nonstationary Reinforcement Learning

Ensuring safety in reinforcement learning under nonstationarity requires anticipating changes in risk before t

センサ/時系列機械学習時系列生成予測テキスト

用途: 生成
難易度: Hard
コスト: High

表形式向き説明可能CPUで試しやすいセンサ/時系列深層学習Transformer回帰テキスト表形式

A Machine Learning Framework for Predicting Restaurant Food Waste to Support Sustainable Food Management

Food waste in the restaurant sector poses a substantial challenge to environmental sustainability and economic

用途: 回帰
難易度: Hard
コスト: Low

Risk-Conditioned Fine-Tuning of Large Language Models

Large Language Models (LLMs) are increasingly deployed in settings where rare but severe harmful generations c

用途: 生成
難易度: Hard
コスト: High

Procedural Graphs: Self-Evolving Execution Structures for LLM Agents

Large language models are increasingly deployed as agents that plan over long horizons and act through externa

用途: 生成
難易度: Hard
コスト: High

Co-Evolving Harnesses and Models: On-Policy Correction Helps Weaker Models Catch Up Where Imitation Fails

Agent harnesses (the system prompt, tool set, execution hooks, and context-management scaffolding around a mod

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Canonical Color as a Lens into Concept Decodability in Vision Encoders and VLMs

Visual encoders construct a representation of the image input for Vision-Language models. How much conceptual,

コンピュータビジョンマルチモーダル画像テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

MeClear: Cooperative Game-Theoretic Attribution and Risk-Aware Memory Clearance for Long-Horizon LLM Agents

Long horizon Large Language Model (LLM) agents rely on external memory systems to preserve user preferences an

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Measuring LLM Sycophancy under Sustained Multi-Turn Pressure

Large language models (LLMs) may abandon correct positions when users push back, exhibiting a failure mode kno

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

説明可能自然言語処理埋め込み・検索検出セグメンテーション画像

GoDeep: Annotation-Free Open-Vocabulary 3D Scene Understanding via Language-Space Lifting

Open vocabulary 3D semantic segmentation methods typically lift CLIP features into 3D. This embeds points in a

用途: 検出
難易度: Hard
コスト: High

Performance of Clinical AI System and Physicians and Frontier Language Models in primary care diagnostics

Clinical AI evaluation should encompass diagnosis and management after adaptive information gathering. We comp

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

The Audit Decides the Verdict: Instrument Effects Rival Demographic Bias in LLM Decision Audits

Whether a language model looks demographically biased can depend on how the audit asks its question. A charita

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

PlannerForge: LLM Agents for Scenario-Based Testing of Motion Planners in Autonomous Driving

Ensuring the safety of autonomous driving is a critical challenge. Scenario-based testing is a systematic proc

深層学習Transformer生成テキスト3D

用途: 生成
難易度: Hard
コスト: High

表形式向き自然言語処理大規模言語モデルテキスト表形式

SQLMorph: Query Mutation and Fine-Grained Metrics for Text-to-SQL Evaluation

Text-to-SQL systems translate natural language queries into executable SQL, democratizing access to structured

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

SkillAdam: Stable and Efficient Skill Evolution for Agents

Agent skills provide a lightweight way to equip frozen language-model agents with domain knowledge and procedu

用途: 技術検証・論文読解補助
難易度: Easy
コスト: Medium

Evaluating and Improving Evidence-Grounded Fact-Checking in LLMs via Multi-Round Evidence Ablation

Automatic fact-checking systems assess the veracity of claims given evidence from relevant documents. Large La

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

OntoKG-EQ: A provenance-grounded, competency-question-governed knowledge graph for auditable analyst querying

Analysts in emerging equity markets keep answering the same questions. Did fundamentals match the market's res

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

API Benchmark Scores Do Not Reliably Transfer to Chatbot Interfaces

Benchmark scores are a central currency in model releases: they inform purchasing decisions, shape public trus

用途: 生成
難易度: Hard
コスト: High

Closing the Consistency Gap: Self-Evolving Agents That Learn to Stay on Course

Large language model (LLM)-powered agents can be accurate on average yet unreliable in production, a discrepan

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Hi-FLoop: Hierarchical State-Feedback Loops for Multi-Timescale World Modeling

Multi-agent traffic simulation seeks diverse, coordinated, and physically realistic futures from maps and obse

強化学習モデルベース生成テキスト

用途: 生成
難易度: Hard
コスト: Medium

Silent Revision: Measuring Undisclosed Change in the Safety Frameworks of Frontier AI Developers

Frontier AI developers publish safety frameworks that commit them to evidencing whether their models are dange

機械学習教師あり学習テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

少数データ向きセンサ/時系列深層学習Transformer予測テキスト時系列

It's All in the Way You Say It: The Role of Information Representation in LLM-Based Glycemic-Event Prediction

Large Language Models (LLMs) are increasingly being investigated for physiological time-series prediction, yet

用途: 予測
難易度: Hard
コスト: High

Benchmark Scores Are Pipeline-Dependent: A Reliability Audit of Cybersecurity LLM Benchmarks

Large language model (LLM) benchmarks are often treated as fixed datasets with stable scores, yet their outcom

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Kairos: A Dataset for Fine-Grained Video-Language Modeling over Space, Time, and Dynamics

Many emerging video language modeling tasks require systems to move beyond clip-level abstraction and model vi

コンピュータビジョン動画認識生成埋め込み画像

用途: 生成
難易度: Hard
コスト: High

品質予測/異常検知深層学習軽量化・量子化検出生成テキスト

CausalChapter: Improving Long-Video Chaptering with Interventional Dependency Modeling

Long-form instructional videos require automatic chaptering to support browsing, navigation, and knowledge acc

用途: 検出
難易度: Hard
コスト: High

X2Streaming-ASR: wait when uncertain, emit when ready for streaming ASR

Streaming automatic speech recognition (ASR) for real-time voice agents and full-duplex dialogue must provide

センサ/時系列自然言語処理RAG分類テキスト音声

用途: 分類
難易度: Hard
コスト: High

CASD: Chunk-Aligned Semantic Distillation for Multi-StageRobot Manipulation

An action chunk can span several stages of a manipulation task, yet a label for its first step describes only

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

CLAMP: Constrained Decoding for Vision-Language Embodied Planning

Embodied planning increasingly relies on vision-language models (VLMs) to translate instructions and visual ob

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Graph-Based Personalized Memory for LLM Agents: Representation, Evolution, Retrieval, and Evaluation

Large Language Model (LLM) agents are evolving from single-session tools toward long-term personal assistants

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

A Three-Tier Persona Vector for Controllable User Simulation in Agentic Evaluation

Evaluating tool-augmented LLM agents requires diverse, realistic user inputs yet most evaluation frameworks us

用途: 生成
難易度: Hard
コスト: High

The Unreliable Progress Bar: Can LLM Agents Reliably Report Task Progress Throughout Execution?

Recent large language models can emit task-progress signals that agent frameworks use to decide whether a task

用途: 生成
難易度: Hard
コスト: High

Do New Attention Mechanisms Actually Fix Attention Sinks at Million-Token Context?

Long context language models now advertise windows of one million tokens, but two habits limit how much of tha

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

AgentGrad: Intervention-guided Prompt Optimization for Multi Agent Systems

Large language model (LLM)-based multi-agent systems (MAS) achieve strong performance by employing specialized

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

BIO-MEMART: Biometric-Aware KV Cache Memory for Multi-User LLM Agents

KV cache is evolving from a serving optimization into an external memory substrate for long-term LLM agents. I

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Same Values, Different Languages? From Multilingual Probing to Steering LLMs Toward Chinese Social Values

As Large Language Models (LLMs) are increasingly integrated into human society, aligning them with pluralistic

表形式向き自然言語処理大規模言語モデルテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

SRPO: Setwise Relative Policy Optimization for Multi-Agent LLMs

Multi-agent large language models solve complex tasks by coordinating several policies in a shared environment

自然言語処理大規模言語モデルテキスト強化学習

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

SequenceO1: End-to-End Ultra-Long (100K) Sequence Modeling in Recommendation with Low-Rank Caching

Modeling long-term user behavior is central to sequential recommendation and billion-scale industrial recommen

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

EvolveScaler: Synthesizing Information-Evolution Contexts via Executable State Machines and Natural-Language Rendering

In persistent interactions, long contexts may encode an evolving process rather than a fixed record: later eve

自然言語処理大規模言語モデル生成異常検知テキスト

用途: 生成
難易度: Hard
コスト: High

Feyospace-v1: How the Cyber Mercury Seven Trained Frontier Cyber Models

Training capable cyber agents is often treated primarily as a problem of model scale, yet open-weight post-tra

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知深層学習軽量化・量子化生成テキストマルチモーダル

FastE: Readout-Triggered Token Compression for LLM Embedding Inference

In this study, we identify depth-dependent prefix redundancy in final-readout LLM embedding models, notably ac

用途: 生成
難易度: Hard
コスト: High

自然言語処理大規模言語モデル画像テキストマルチモーダル

Towards Embodied Air-Ground Cooperative Object Search: Benchmark, Dataset and Agentic Method

Air-Ground Object Search (AGOS) in urban environments is a challenging embodied task, which requires an Unmann

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

コンピュータビジョンセグメンテーション分類検出画像

Segment Any Motion with Radar: Robust Multimodal Moving-Object Segmentation and Tracking

Moving-object perception must decide which image regions correspond to real motion and keep every instance ide

用途: 分類
難易度: Hard
コスト: High

AttnCompress: Dynamic Attention-Guided Trajectory Compression for Software Engineering Agents

The transition from human-centric assistance to Autonomous Software Engineering (ASE) agents has enabled the r

深層学習軽量化・量子化セグメンテーションテキスト

用途: セグメンテーション
難易度: Hard
コスト: Medium

FPicker: Topology-Guided Evolution for Filament Tracing in Low-SNR Microscopy

Automating filament tracing in Cryo-Electron Microscopy (Cryo-EM) is essential for 3D helical reconstruction b

深層学習Transformer検出画像テキスト

用途: 検出
難易度: Hard
コスト: High

少数データ向き自然言語処理大規模言語モデル生成テキスト

LEBGen: An LLM-Enhanced Bayesian Network Framework for Few-Shot Travel Survey Data Generation

Travel survey data are essential for transportation planning and travel behavior analysis, yet collecting larg

用途: 生成
難易度: Hard
コスト: High

What Eviction Destroys: A Restore-Counterfactual Audit of Forgetting in Agent Memory

Agent memory systems must discard stored information when their history exceeds a fixed token budget. Existing

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

MemForest: Efficient Agent Memory Management via EventTree Partitioning and Progressive Merging

Agent memory systems have demonstrated significant potential in long-term dialogue, personalized assistants, a

深層学習軽量化・量子化テキスト動画マルチモーダル

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Evidence-Aligned Entity Verification for Hallucination Detection in Retrieval-Augmented Generation

Hallucination detection is crucial for large language models (LLMs), as hallucinated content creates significa

自然言語処理大規模言語モデル検出生成テキスト

用途: 検出
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデル画像テキスト

ACEA: An Adversarial Co-Evolution Arena for Head-to-Head Red-Team and Blue-Team LLM Testing

Automated red-team attacks and blue-team defenses for large language models (LLMs) are advancing quickly. Howe

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

MI向き自然言語処理ファインチューニングテキストマルチモーダル

CircuTutor: Transforming Static Circuit Problems into Intelligent and Dynamic Tutoring

Learning direct current circuit concepts requires learners to connect invisible physical quantities, such as c

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

CS-CLIP: Compositional Scene Graph-guided CLIP for Robust Compositional Reasoning

Vision-language models (VLMs) demonstrate strong performance across compositional reasoning benchmarks, which

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

表形式向き深層学習グラフニューラルネット分類テキスト

TTGBench: Benchmarking Topological Evolution and Semantic Drift in Text-attributed Temporal Graphs

Temporal graph learning models the evolution of dynamic systems, where both structural interactions and semant

用途: 分類
難易度: Hard
コスト: High

3DWay: Generalizing Robot Manipulation via 3D Consistent Waypoints

Intermediate representations are key to bridging the modality gap between generalizable manipulation policies

コンピュータビジョン3D・点群画像テキスト3D

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Do Dynamic Routers Need Memory? HeRo: History-Aware Routing for Efficient LLM Inference

Dynamic layer routing reduces the inference cost of Large Language Models (LLMs) by learning to skip layers fo

深層学習軽量化・量子化生成テキスト

用途: 生成
難易度: Hard
コスト: High

Bridging the Semantic-Utility Gap in Multimodal RAG via Generator-in-the-Loop Alignment

Vision-language models (VLMs) augmented with retrieval-augmented generation (RAG) benefit from access to exter

自然言語処理RAG生成画像テキスト

用途: 生成
難易度: Hard
コスト: High

Less Is Personal: Learning Minimal Sufficient User Profiles for Personalized Language Models

Retrieval-augmented personalization enables large language models to produce more accurate and preference-alig

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

強化学習方策勾配 (PPO / A3C)生成テキスト

Safe Harness Self-Evolution: A Theoretical Analysis of Feasibility and Limits

Harness self-evolution is the process by which an agent modifies its prompts, tools, code, or orchestration in

用途: 生成
難易度: Hard
コスト: Medium

OntologyBench: Can Dense Retrieval Satisfy Structured Biomedical Constraints?

We introduce OntologyBench, a tiered biomedical retrieval benchmark comprising 471,854 training and 125,744 ev

自然言語処理大規模言語モデル検索テキスト

用途: 検索
難易度: Hard
コスト: High

説明可能品質予測/異常検知自然言語処理大規模言語モデルテキスト

Key Path Identification for Resolving Knowledge Conflicts via SAE-based Steering

Sparse autoencoder (SAE)-based steering has been widely used to address knowledge conflicts by guiding LLMs to

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Dual-Layer Semantic-Spatial Belief Mapping for Aerial Object Goal Navigation

Aerial Object Goal Navigation (ObjectNav) requires an unmanned aerial vehicle (UAV) to locate a described targ

自然言語処理RAG画像テキストマルチモーダル

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

SWE-Bench Pro Verified: A Reliable Benchmark for Software Engineering Agents

SWE-Bench Pro has emerged as a standard benchmark for evaluating software engineering agents on challenging re

品質予測/異常検知機械学習教師あり学習テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

IGT @ FinMMEval 2026 Task 2: Question-Type Prompting with Targeted Extraction for Multilingual Financial QA

We present the IGT system for PolyFiQA Task 2 of the FinMMEval Lab at CLEF 2026, a multilingual financial ques

自然言語処理RAG生成QAテキスト

用途: 生成
難易度: Hard
コスト: Low

Router Prior Bias: Preserving Base Routing Structure in MoE Post-Training

Mixture-of-Experts (MoE) pretraining relies on an auxiliary load-balancing loss (LBL) to drive per-expert util

自然言語処理ファインチューニングテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Artificial Intelligence-Assisted Digital Inventory of Cultural Heritage & Traditional Knowledge: Case for Indonesian Open Digital Library of Culture

The Indonesian Digital Library of Culture (Perpustakaan Digital Budaya Indonesia, PDBI; budaya-indonesia.org)

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

CIVI: A Framework for Diagnosing Search Agent Failures in Civic Information

Large Language Models are increasingly deployed in public-sector settings, where incorrect guidance can cause

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

説明可能品質予測/異常検知自然言語処理大規模言語モデル生成テキスト

Automated Design of Inventory Policy with Large Language Models: An Exploratory Study

Firms making inventory decisions have access to operational data, optimization tools, and large language model

用途: 生成
難易度: Hard
コスト: High

品質予測/異常検知コンピュータビジョンマルチモーダル生成画像テキスト

Studying Image Tokenizers as Visual Languages in Unified Multimodal Models

Image tokenizers define the ``visual language'' of unified multimodal models, yet are commonly studied through

用途: 生成
難易度: Hard
コスト: High

It's Not RoPE that Creates Sinks: The Role of Self-Concentration and Value-Non-Mixing in Attention

Large Language Models (LLMs) often exhibit "Attention Sink" (AS) and the accompanying "Massive Activations" (M

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

ToolLoop: Closed-Loop Tool-Use Data Synthesis via Decomposed Generation and Dynamic Self-Feedback

High-quality tool-use data is critical for training language models to interact effectively with external tool

用途: 生成
難易度: Hard
コスト: High

センサ/時系列深層学習Transformer分類生成テキスト

AuK Technical Report: An Open-Source Foundational Model for Speech Generation and Editing

We introduce AuK, an open-source foundational model that unifies speech generation and editing through a commo

用途: 分類
難易度: Hard
コスト: High

コンピュータビジョンセグメンテーションQAテキスト

When Models Defer to Wrong Answers: A Robustness Audit of Source-Attributed Cues in Multiple-Choice QA

Language models often receive a question together with a claim about what another source answered. We audit wh

用途: QA
難易度: Hard
コスト: Medium

Experience Funnel: A State-Policy Alternating Loop for Self-Evolving Agents

Autonomous agents powered by large language models (LLMs) continuously accumulate experience through interacti

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Evolution of Multimodal Question Answering: From Modality-Adaptive Extraction to Unified Language Representation

The rapid growth of multimodal data has intensified the need for question answering (QA) systems capable of re

コンピュータビジョンマルチモーダル生成QA画像

用途: 生成
難易度: Hard
コスト: High

Q2D-Web: A Large-Scale Benchmark for Retrieval in Agentic RAG Systems

Evaluating first-stage retrievers in large-scale production RAG requires a benchmark that pairs a large-scale

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

The Rater Ising-Potts Model with LLM-Derived Weights: An Application to Multi-Category Scoring Reliability

The Ising model is extended to the Potts model for multinomial data. We introduce a Rater Ising-Potts model th

説明可能自然言語処理大規模言語モデル分類テキスト

用途: 分類
難易度: Hard
コスト: High

Improving Term Evaluation in Machine Translation: Variation Matters

Terminology evaluation in machine translation (MT) usually assumes a single correct target form per source ter

自然言語処理翻訳テキスト

用途: 翻訳
難易度: Hard
コスト: High

Record Grouping Controls Evidence Weight in Language Models

Retrieved records are presentation units; a supplied partition determines which records enter a language model

自然言語処理生成テキスト

用途: 生成
難易度: Hard
コスト: Medium

Global Divergence, Local Convergence: Representation Geometry in SSMs and Transformers

Recent state-space models (SSMs) such as Mamba achieve language modeling performance comparable to transformer

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

When Victorian Becomes a Prompt: Literary Periodization as a Generative Constraint in 100 AI-Generated Novels

Generative AI inverts the typical periodization of literary history: the periodizing tag Victorian can now com

用途: 生成
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデル検出生成テキスト

Combating Instruction Conflict via Energy-Driven Latent Conflict Detection

Large Language Models (LLMs) are increasingly deployed with hierarchical instructions, yet they remain vulnera

用途: 検出
難易度: Hard
コスト: High

Navigating the digital spectrum: Assessing political bias, stability, and downstream fairness in Large Language Models

Large Language Models are increasingly deployed as information intermediaries, yet measuring their political b

深層学習Transformer検出テキスト音声

用途: 検出
難易度: Hard
コスト: High

Dynamics of meaning: Towards the Evaluation of Diachronic Semantic Change in Sinhala

Tracking semantic change in low-resource languages across extensive historical timelines presents significant

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

Which Forms of Caregiver Feedback Support Grammar Learning? A Reinforcement-Learning Study of Child-Like Language Models

Social interaction is central to children's language learning, but the effects of different forms of caregiver

自然言語処理大規模言語モデル生成テキスト強化学習

用途: 生成
難易度: Hard
コスト: High

Detecting Authorship in Political Texts with Inductive Stylometry

Political texts are rarely authored by the nominal speaker alone. Tweets, speeches, reports, and official stat

自然言語処理テキスト音声

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Compositional Multilingual and Behavioral Attribute Steering

This study examines the compositionality of steering vectors for language and behavioral control in large lang

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

センサ/時系列自然言語処理大規模言語モデル検出生成QA

From Coordinates to Candidate Regions: Temporal Change Localization via Region Selection in Remote Sensing Multimodal LLMs

Remote sensing multimodal large language models (RS-MLLMs) have advanced scene understanding and visual questi

用途: 検出
難易度: Hard
コスト: High

Structural Jailbreaks Generalize but Do Not Compound: A cross-provider and multilingual study of Involuntary In-Context Learning

Aligned language models fail under two independent pressures: the structural jailbreak class recently formaliz

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Reading a Legal Question Word by Word: Embedding Trajectories of 2,144 Vietnamese Legal Headlines

A dense retriever encodes a question as one vector, but the question arrives one word at a time. We read 2,144

自然言語処理埋め込み・検索テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

SentryLine: Evidence-Grounded Question Answering over Evolving Documents in Oncology Care

Oncology care operates at constant pressure of absorbing rapidly evolving evidence base in biomedicine. The Am

自然言語処理大規模言語モデル検出生成QA

用途: 検出
難易度: Hard
コスト: High

NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness

Recursive self-improvement (RSI) requires a concrete mechanism through which an AI system observes its capabil

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデル要約テキスト音声

EviSI: An Evaluation Agent for Simultaneous Interpreting

Simultaneous speech-to-speech translation requires understanding, translation and spoken delivery while the so

用途: 要約
難易度: Hard
コスト: High

Snugi-AI-v2 @ eRisk 2026 Task 2: Early Depression Detection via a Learned Stopping Policy with Sustained Confidence Gate

We describe the Snugi-AI-v2 submission to eRisk 2026 Task 2, the second edition of contextualized early depres

深層学習Transformer分類検出テキスト

用途: 分類
難易度: Hard
コスト: High

品質予測/異常検知深層学習Transformerテキスト強化学習

When Metrics Reward the Worst Translations: Internalizing Cultural Reasoning for Social Media Translation Evaluation

Automatic translation quality metrics trained on general-domain corpora systematically fail on social media co

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデルテキスト音声

ConversationalVoice: Full-Duplex Speech Data from Real Conversations through Source-Faithful Reconstruction and Conversation-Grounded Expansion

Full-duplex speech models require training data that preserves turn-taking, overlap, interruption, and backcha

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Jacap: Robust KV Cache Eviction via Jacobian-Based Nonlinear Information Capacity Preservation

Key-value (KV) cache eviction is essential for scaling long-context inference in Large Language Models. Howeve

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Popular Knowledge Propagates More Errors in LLM Knowledge Updating

Updating a language model's knowledge through fine-tuning is essential for keeping its outputs current, yet ca

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

SyncWorld: Visual Calibration Enables World Models as Zero-Shot Simulators

World models are increasingly used as policy-in-the-loop imagination environments, where reliable rollouts req

自然言語処理RAG画像テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Rethinking Learned Occupancy in Autonomous Active Mapping with Observation-Gated Filtering

Autonomous 3D active mapping requires a space robot to choose where to sense while building the geometry neede

センサ/時系列自然言語処理RAG検出テキスト3D

用途: 検出
難易度: Hard
コスト: High

Spheriverse: 3D Scene Understanding from Spherical Observations in the Wild

Spherical observations provide global visual context for 3D scene understanding. However, visual information i

コンピュータビジョン物体検出検出画像テキスト

用途: 検出
難易度: Hard
コスト: High

Concentrate After Imagination: Text-Conditioned Evidence Grounding for Partially Relevant Video Retrieval

Partially Relevant Video Retrieval (PRVR) retrieves untrimmed videos when queries describe only short moments.

深層学習Transformerテキスト動画

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

FRAME: Factored Retrieval via Attribute Readouts for Object-Centric Scene Memory

Language-guided robots need persistent scene memories to follow instructions, revisit objects, and resolve ref

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

自然言語処理大規模言語モデル検出セグメンテーションテキスト

SeGDeP: Semantic- and Geometric-Aware Decoupled Prompts for Reasoning Segmentation

Reasoning segmentation converts an implicit linguistic conclusion into a precise mask, requiring both semantic

用途: 検出
難易度: Hard
コスト: High

DSE-VTG: Dual-Side Enhancement for Training-Free Video Temporal Grounding

Text-guided Video Temporal Grounding (VTG) aims to localize the relevant segments in an untrimmed video based

深層学習軽量化・量子化画像テキスト動画

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知コンピュータビジョン物体検出検出画像テキスト

FIRE3D: Feed-forward Interactive 3D Scene Reconstruction Within A Minute

We present FIRE3D, a unified framework that takes a single RGB image or casual RGB video and transforms it int

用途: 検出
難易度: Hard
コスト: High

Beyond Gait: Person Identification from Millimeter-Wave Point Clouds Across Activities of Daily Living

Person identification from millimeter-wave (mmWave) point clouds has mainly relied on gait. Indoor walking, ho

深層学習Transformer分類テキスト3D

用途: 分類
難易度: Hard
コスト: High

自然言語処理プロンプトエンジニアリング生成セグメンテーション画像

MorphoOrgaAgent: A Foundation-Model-Based Multi-Agent System for Autonomous Organoid Analysis

Organoids are three-dimensional tissue models whose morphology provides important insights into tumor developm

用途: 生成
難易度: Easy
コスト: Medium

品質予測/異常検知深層学習Transformerセグメンテーション画像テキスト

CoordFormer: Give Me Any Coordinates and I Will Give You Labels

Semantic segmentation on very-high-resolution images remains challenging due to the high computational cost an

用途: セグメンテーション
難易度: Hard
コスト: Low

センサ/時系列深層学習軽量化・量子化検出画像テキスト

MFVINS: Multiple Fisheye Camera-Based Visual Inertial System

A simultaneous localization and mapping (SLAM) method using a monocular camera and a low-cost inertial measure

用途: 検出
難易度: Hard
コスト: High

GOLF: Global Observation with Local Focus for Calibration-Aware Stereo Interaction Field Estimation

We present GOLF, the first-place solution to the SHOW3D Interaction Field Estimation Challenge at HANDS@ECCV 2

深層学習Transformerテキスト3D

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

深層学習TransformerQAテキストマルチモーダル

STSG-VQA: Evidence-Grounded Temporal Question Answering from Surgical Spatio-Temporal Scene Graphs

Despite recent advances in surgical vision-language models (VLMs), temporal reasoning remains limited because

用途: QA
難易度: Hard
コスト: High

ReMoMask-2: Latent Retrieval-Augmented Masked Motion Generation

Text-to-motion (T2M) generation maps natural language to human joint movements, aiding gaming, VR, and robotic

深層学習Transformer生成テキスト3D

用途: 生成
難易度: Hard
コスト: High

VeriScene: Reconstructing Crime Scenes from Legal Evidence via World-Model Agent

World models take multimodal inputs like text, photos, and diagrams to generate dynamic scenes in accordance w

自然言語処理大規模言語モデル画像テキスト動画

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Do Input-Level Defenses Transfer to Observation-Level Attacks on VideoLLMs?

Video Large Language Models (VideoLLMs) are increasingly deployed in safety-critical applications such as cont

深層学習軽量化・量子化検出画像テキスト

用途: 検出
難易度: Hard
コスト: High

Supervised Cross-Modal Feature Alignment for Zero-Wearable Freezing of Gait Detection in Parkinsonism

Objective assessment of Freezing of Gait (FoG) in Parkinson's disease (PD) relies predominantly on wearable In

センサ/時系列深層学習軽量化・量子化分類検出画像

用途: 分類
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデル分類検出生成

From Glance to Scrutiny: Progressive Distortion Reasoning for Fine-Grained Image Quality Assessment

Multi-modal large language models (MLLMs) have demonstrated significant potential in image quality assessment

用途: 分類
難易度: Hard
コスト: High

深層学習Transformerセグメンテーションテキスト3D

TRIUNE-Net: Harmonizing Scale, Shape, and Efficiency in Pancreatic Tumor Segmentation

Pancreatic tumor segmentation in 3D CT volumes is challenged by extreme scale variability across both the panc

用途: セグメンテーション
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデル検出生成画像

Human-Centric Image Captioning with Subject-Centered Spatial Understanding

While multimodal large language models (MLLMs) achieve remarkable performance on generic image captioning, the

用途: 検出
難易度: Hard
コスト: High

深層学習Transformerセグメンテーション画像テキスト

MARS-CLIP: Multi-Resolution and Attention Refined Zero-Shot Image Segmentation

Contrastive Language-Image Pre-training (CLIP) has demonstrated impressive capabilities in zero-shot transfer

用途: セグメンテーション
難易度: Hard
コスト: High

コンピュータビジョンセグメンテーション生成画像テキスト

Dreaming in Flow: Generative Grounding Feedback for Self-Evolving Unified Multimodal Models

Unified multimodal models integrate visual understanding and generation within a single network, yet the two c

用途: 生成
難易度: Hard
コスト: High

Drive by Hindsight and Foresight: Tool-Grounded Synergistic Reasoning over Hierarchical Memory for Autonomous Driving

VLMs have shown promise for autonomous driving, yet still suffer from hallucination, weak spatio-temporal perc

用途: 生成
難易度: Hard
コスト: High

PhysFlow: Physics-Aware Optical Flow for Motion Controllable Video Generation

Video generation models have recently attracted substantial attention for their ability to generate visually c

自然言語処理RAG生成画像テキスト

用途: 生成
難易度: Hard
コスト: High

Observe Before You Alert: Adaptive Driver Alerting with Vision-Language Models

Driver alerting from dashcam video requires sequential decision-making under partial observability: a system m

深層学習軽量化・量子化生成テキスト動画

用途: 生成
難易度: Hard
コスト: High

コンピュータビジョンセグメンテーション予測テキスト動画

DriveMotion: A Large-Scale Multi-Source Benchmark for Driver Motion Sequence Modeling and Forecasting

Driver motion can provide cues to ongoing behavior, attention, and near-term driving intent. However, most exi

用途: 予測
難易度: Hard
コスト: High

Remotely Detectable Keyed Communication through Motion

Messages from electronic devices are conventionally received as text, audio, or radio signals. But robots move

センサ/時系列強化学習テキスト音声動画

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

FOCI Policy: Focus on Object-Centric Interactions for Relational Manipulation Policies

Object-centric manipulation policies improve generalization by modeling object motion instead of directly pred

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

説明可能コンピュータビジョンセグメンテーションテキスト

HiBRIDGE: A Hierarchical Bayesian Neural Network Framework for Interpretable Dialogue Management in Group-Robot Interaction

In multi-party human-robot interaction, a robot must continuously decide whom to address and what to say to pa

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Estimating Semantic Ambiguity via Gaussian Context Distributions for VLM-Driven Traversability Analysis

Autonomous navigation in unstructured environments requires robust scene understanding, yet Vision-Language Mo

コンピュータビジョンマルチモーダル画像テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

自然言語処理大規模言語モデルテキスト動画マルチモーダル

Safe Task Planning with Long-Term Graph Memory for Embodied Agents

Large language models (LLMs) and vision-language models (VLMs) have significantly advanced zero-shot task plan

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Localized Visual Feature Aggregation via Focus Pooling for Visuomotor Policies

Focusing on spatially localized, control-relevant visual cues has been shown to improve data efficiency in vis

深層学習CNN画像テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

EvoNav-Bench: Benchmarking Lifelong Navigation in Evolving Environments

Lifelong navigation (LN) requires an embodied agent to solve a sequence of navigation subtasks in the same env

深層学習Transformer画像テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Bridging Language and Physics: Automated Design of Continuum Robots with Large Language Models

Large language models (LLMs) have recently emerged as a promising tool for automating robot design from high-l

自然言語処理大規模言語モデル生成テキスト強化学習

用途: 生成
難易度: Hard
コスト: High

センサ/時系列深層学習Transformerテキスト

TacClip: a clip-on sensor measures dynamic contact forces without covering the fingerpads

TacClip is a minimally encumbering wearable device for recording fingertip deformation caused by contact force

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

OmniNav: Robust Long-Horizon Target Navigation in Dynamic Environments

Long-horizon target navigation requires a robot to sustain task execution across evolving observations, decisi

自然言語処理ファインチューニングテキスト3D

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知強化学習方策勾配 (PPO / A3C)テキスト

Approval-Based Multiwinner Voting with Candidate Qualities

We initiate the study of a new model of approval-based multiwinner voting in which each candidate carries an e

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Stable Voting Rules on the Edge of Optimal Metric Distortion

We prove the existence of a randomized voting rule with metric distortion at most $2.13713$, within $0.025$ of

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

MI-PEFT: Mixture-of-Experts Integrated Parameter-Efficient Fine-Tuning Protein Language Models Improves Acidophilic Proteins Classification

Acidophilic proteins that remain stable and functional under highly acidic conditions, are important for indus

深層学習軽量化・量子化分類テキスト

用途: 分類
難易度: Hard
コスト: High

Bayesian Matrix-Valued Graphs for Context-Dependent Multivariate Relationships

Many scientific graphs attach several variables to each node, so a single scalar edge weight cannot describe d

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Flexible Motion Generation from Language and Style References

We introduce FlexMoGen, a novel framework for flexible human motion synthesis conditioned on both natural lang

用途: 生成
難易度: Hard
コスト: High

品質予測/異常検知深層学習軽量化・量子化分類回帰テキスト

Automated Chest CT Protocol Selection via Large Language Model Derived Text Embeddings from Imaging Request Text

Purpose: Accurate CT protocol selection is critical for diagnostic quality and patient safety, yet the current

用途: 分類
難易度: Hard
コスト: High

品質予測/異常検知コンピュータビジョンセグメンテーション回帰テキスト教師あり

Semi-Supervised Learning under Spatially Biased Sampling

Standard semi-supervised learning (SSL) typically relies on labelled and unlabelled data sharing a common marg

用途: 回帰
難易度: Hard
コスト: Medium

MetaKV: Adaptive KV Cache Compression for Constrained LLM Inference

Key--value (KV) cache compression is an effective way to reduce the memory overhead of large language model (L

深層学習軽量化・量子化QAテキスト

用途: QA
難易度: Hard
コスト: High

表形式向き自然言語処理プロンプトエンジニアリングテキスト表形式

Streaming Hierarchical Inference with Tabular Foundation Models

Tabular Foundation Models (TFMs) have recently demonstrated strong predictive performance through in-context l

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

Beliefs and Behavior in Language Models

There is significant uncertainty about whether abstractions like beliefs or desires usefully describe the beha

説明可能自然言語処理大規模言語モデルテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

LLM Layers Immediately Correct Each Other

Recent methods in language model interpretability employ techniques such as sparse autoencoders to decompose r

説明可能深層学習Transformerテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

InfluenceField: A Differentiable Field with Interventionally Identifiable Causal Structure for Multimodal World Modeling

Multimodal large language models often capture visual-linguistic correlations but struggle to predict how loca

深層学習軽量化・量子化画像テキストマルチモーダル

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

A*-Thought-V2: Efficient Latent Reasoning via Geometric Dynamics of LLM

Chain-of-Thought (CoT) improves the reasoning ability of Large Language Models (LLMs) but incurs substantial c

深層学習軽量化・量子化テキスト3D

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Kalman Delta Networks: Uncertainty-aware Associative Memory

Linear attention is increasingly used in frontier language models for efficient long-context inference and con

説明可能深層学習RNN / LSTMテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

You Can't Prefer Emotions You Don't Sample: Intensity Undershoot in DPO-Tuned LLMs

Ask a language model to respond "very excitedly," and its output is typically only mildly more energetic. We q

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

センサ/時系列品質予測/異常検知深層学習Transformerテキスト時系列

Decomposition-Guided Diffusion Language Models for Inertial Confinement Fusion Prediction

Inertial confinement fusion (ICF) is a leading pathway toward clean energy, but each shot at the National Igni

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

説明可能自然言語処理プロンプトエンジニアリングテキスト

On the Recall Scaling Laws in Mamba: A Theoretical and Mechanistic Study via Hashing

Associative Recall (AR) is the cognitive ability to learn and retrieve links between items in memory. In NLP,

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

MpSub: A Momentum $p$-Dimensional Subspace Trust-Region Method for Derivative-Free Fine-Tuning of Large Language Models

Full-parameter fine-tuning of large language models has substantial memory costs because backpropagation store

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Accuracy is Not Enough: A Divergence-Based Approach to Evaluate Fidelity Loss in Quantized LLMs

Deployment of Large Language Models (LLMs) on memory-constrained edge devices relies heavily on aggressive pos

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

少数データ向きCPUで試しやすい条件最適化コンピュータビジョンセグメンテーション生成テキスト

The Art of Hierarchical Competing Patterns: Gaussian Process Optimization of Hyphenation

Hyphenation patterns remain a compact and widely deployed solution for word breaking in typesetting systems, t

用途: 生成
難易度: Hard
コスト: High

I Don't Miss You, but I Do: Self-Explanation Faithfulness of Modality Missingness in Vision-Language Models

Vision-language models are increasingly used in settings where some input modalities may be unavailable, yet w

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

EigenLI: Spectral Approximations to Late Interaction

Late-interaction models such as ColBERT achieve strong effectiveness by representing each document with many t

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

Heat Kernel Textures: the Geodesic Gaussians That Do Not Splat

3D Gaussian Splatting has recently revolutionised novel view synthesis as well as many other 3D vision methods

コンピュータビジョン3D・点群生成画像テキスト

用途: 生成
難易度: Hard
コスト: High

品質予測/異常検知深層学習軽量化・量子化検出生成異常検知

Parser-Free VLM Verification for Federated Weakly Supervised Video Anomaly Detection

How can vision-language models help video anomaly detection (VAD) when surveillance data remain distributed, w

用途: 検出
難易度: Hard
コスト: High

深層学習Transformer分類埋め込みテキスト

RAFM-SER++: A Lightweight Multimodal Emotion Recognition Framework for Real-Time Behavioral Monitoring in Surveillance Systems

Recent multimodal Speech Emotion Recognition (SER) systems achieve high accuracy through interaction-heavy cro

用途: 分類
難易度: Hard
コスト: High

Think Wider: Mitigating Latent Rank Collapse in Implicit Chain-of-Thought Reasoning

Chain-of-thought (CoT) reasoning improves the reasoning ability of large language models by introducing interm

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Dense Structural Compression of Transformers via Gauge-Correct Channel Removal

Inference energy per token drives the cost and carbon footprint of deployed transformers. It is dominated by d

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

LLMs for Social Network Modeling: From Network Generation to Dynamic Processes

Large language models (LLMs) are rapidly emerging as a new paradigm for modeling social networks by representi

用途: 生成
難易度: Hard
コスト: High

Eliciting Self-Verification in Multimodal Reasoning Agents with Reinforcement Learning

Reasoning agents increasingly rely on external tools such as web search to answer complex queries. Reinforceme

自然言語処理RAGQA画像テキスト

用途: QA
難易度: Hard
コスト: High

A Layered Analysis of Disagreement And Answer Quality in Multi-Agent LLM Debate

Multi-agent debate, in which several LLMs exchange arguments before answering, is widely assumed to improve an

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Sparks of In Silico Cognitive Science: Theories from Simulated Data Can Generalize to Humans

Behavioral foundation models have been proposed as stand-ins for human participants across settings, but it is

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

表形式向き自然言語処理大規模言語モデル回帰テキスト

CausalVerify: An Execution-Grounded Benchmark for LLM Causal Inference Workflows

Existing causal-inference benchmarks for LLMs mostly score method descriptions or whether generated code runs,

用途: 回帰
難易度: Hard
コスト: High

深層学習TransformerセグメンテーションQA画像

TDDN: Text-aligned Diffused DINO Network for Puzzle Understanding

Structured visual reasoning, such as image puzzles, demands fine-grained visual perception, an ability current

用途: セグメンテーション
難易度: Hard
コスト: High

Do Large Language Models Know What They Don't Know II? A Fully Behavioral, Non-Cognitive Measure of Epistemic Honesty

Large Language Models (LLMs) are frequently confident, eloquent, and well versed. A natural question arises: d

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデル分類画像テキスト

SAFIRE: Safety-Critical Benchmark for Fine-grained Fire and Smoke Understanding in Multimodal LLMs

Multimodal Large Language Models (MLLMs) show strong progress on vision-language tasks, yet their reliability

用途: 分類
難易度: Hard
コスト: High

説明可能深層学習Transformer生成画像テキスト

VoT: Vision-of-Thought for Unified Multimodal Representation Alignment

Current text-to-image systems typically employ a "text encoder plus diffusion decoder" paradigm, in which text

用途: 生成
難易度: Hard
コスト: High

Climate-ModernBERT: Revisiting Corpus Composition for Domain-Adaptive Continued Pretraining

Natural Language Processing (NLP) in the climate domain requires models to process heterogeneous text sources,

MI向き深層学習Transformerテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

xDailyBench: Benchmarking LLMs on Professional Consultation for Real-Life Problems

Large language models (LLMs) are increasingly used for everyday assistance, yet existing benchmarks only parti

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

説明可能強化学習方策勾配 (PPO / A3C)テキスト

Bag of Tricks or Bag of Myths? Reducing Modeling Complexity with Task Knowledge in Explainable Suicide Risk Assessment

Assessing suicide risk from social media text is a small-data, high-stakes setting requiring not only severity

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Do AI Coding Assistants Check Before They Install? A Pre-Registered Demand-Side Audit of Trust Signals in the Research Software Supply Chain

AI coding assistants now select, install, and configure software, and attackers have exploited that position t

MI向き機械学習教師あり学習テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

TFTrack: A Template-Free Framework for Efficient 3D Point Cloud Tracking

LiDAR-based 3D Single Object Tracking (3D SOT) is critical for robotic perception and navigation and aims to l

深層学習軽量化・量子化検出テキスト3D

用途: 検出
難易度: Hard
コスト: High

コンピュータビジョンセグメンテーション分類検出生成

A radiographic world model for clinical reasoning and evidence generation

Medical imaging artificial intelligence (AI) is commonly developed as separate mappings from radiographs to di

用途: 分類
難易度: Hard
コスト: High

品質予測/異常検知深層学習軽量化・量子化分類テキスト

Fine PT-PT Web: A High-Quality 41 Billion Tokens Data Collection of the European Portuguese Web

Curating Web corpora for regional language variants like European Portuguese (PT-PT) is heavily bottlenecked b

用途: 分類
難易度: Hard
コスト: High

From Citations to Contributions: LLM-Assisted Credit Scoring of Research Articles

Citation-based measures of scientific influence typically treat citations as uniform signals, ignoring the dif

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Aegix Pulse: A Traceable Three-Stage Architecture for Personalized Content Generation and Context-Preserving Revision

Production content-generation systems must integrate a user's immediate task, long-term brand identity, histor

用途: 生成
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデル生成テキスト

Noēsis: Deterministic-First Retrieval with Two-Tier Context Hydration for Factuality-Critical Queries on Small Local Models

A wrong number is worse than no answer. Across factuality-critical domains -- audience metrics, scheduling and

用途: 生成
難易度: Hard
コスト: High

How AI Models Manage Epistemic Authority: A Taxonomy and Comparative Analysis of Responses to User Disagreement

Large language models are increasingly used as sources of advice and information, including in high-stakes set

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Open Tabular Insight Extraction: Where Do We Stand, and Where Should We Go?

Democratizing access to the knowledge held in large corpora of tables such as data lakes is emerging as a cent

表形式向き自然言語処理RAGQA検索テキスト

用途: QA
難易度: Hard
コスト: Low

ObGynLongBench: Revealing the Evidence-to-EHR Gap in Longitudinal EHR Decision-Making

The application of large language models (LLMs) to personalized medical assistants has garnered growing intere

自然言語処理大規模言語モデルQAテキスト

用途: QA
難易度: Hard
コスト: High

説明可能深層学習Transformer分類埋め込みテキスト

Mapping the Emerging Social Science of Large Language Models

Large language models (LLMs) increasingly shape communication, learning, work, creativity, and decision-making

用途: 分類
難易度: Hard
コスト: High

コンピュータビジョンマルチモーダル画像テキスト動画

Large-Scale User Behavior Analysis in Multimodal AI-Assisted Manual Task Execution

Conversational Task Assistants (CTAs) are multimodal dialogue systems that support users in complex real-world

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

説明可能コンピュータビジョンマルチモーダル分類生成画像

BanglaMemeX: Advancing Cultural Metaphoric Image Interpretation in Bangla with a Multimodal Explainable Dataset

Vision Language Models have achieved strong performance on multimodal benchmarks, yet their ability to reason

用途: 分類
難易度: Hard
コスト: High

MeRoTune: RoPE-Safe Merging with a Tunable Dial

When you merge two fine-tuned models from the same base checkpoint by simply averaging their weights, you impl

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Reasoning Beyond Transcription: Audio Language Models on Child Stuttering Speech

Child speech differs from adult speech in acoustics, prosody, and linguistic structures. Speech disfluencies (

自然言語処理大規模言語モデル要約テキスト音声

用途: 要約
難易度: Hard
コスト: High

Deadline-Aware Adaptive Prefill Chunking for Efficient Large Language Model Serving

Continuous batching improves large language model (LLM) serving throughput, but long prompt prefills can delay

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知深層学習Transformer生成テキスト

Does Syntax Matter? A Graph-Augmented Variational Topic Model for Computational Social Sciences

Topic modeling is widely used in computational social sciences to identify latent themes in large text corpora

用途: 生成
難易度: Hard
コスト: Low

LLM Agents as Computational Typologists

Linguistic typology relies on expert analysis of reference grammars across languages, making large-scale cross

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Signed Rescue Routing: Harm-Aware Cascades for Efficient LLM Inference

Large language model (LLM) cascades answer easy requests with a small model and escalate selected requests to

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Replicating a Disjoint-Set Union Experiment over Various Notions of Micro Units to assess Translation Effort

The paper describes a replication experiment to assess "the distribution of editing procedures across micro an

自然言語処理テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

LLM Forensics: Where Do Backdoors Hide? Localizing and Controlling Trigger Mechanisms with Sparse Autoencoders

Even though backdoors in LLMs have been a growing concern, their inner workings are still under heavy scrutiny

深層学習Transformer検出生成テキスト

用途: 検出
難易度: Hard
コスト: High

From Echo Chambers to Epistemic Monoculture: Large Language Models Present Temporally Contingent Partisan Alignments as Knowledge

Large language models (LLMs) are rapidly becoming an interface between citizens and political information. The

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Translation Indeterminacy and the Distributional Fallacy

Large language models (LLMs) are commonly associated with the distributional hypothesis, according to which (1

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

DeepTable: Structural Attention Biases and Tree Path Encoding for Hierarchical Table Understanding

Large language models (LLMs) have demonstrated strong performance in table understanding. However, they typica

深層学習軽量化・量子化QAテキスト

用途: QA
難易度: Hard
コスト: High

Perspectives on Cross-Lingual Consistency in LLMs for Medical Questions

Should multilingual LLMs answer medical questions consistently across input languages, or adapt responses to c

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Syntactic Patterns and Stylistic Functions in Narrative Prose: A Rule-Based and Machine-Learning Approach

This paper presents a small-scale quantitative experiment that links syntactic structure to stylistic function

センサ/時系列機械学習時系列分類テキスト音声

用途: 分類
難易度: Hard
コスト: Low

Same Problem, Different Field: Cross-Domain Solution Import via Domain-Stripped Computational Fingerprints

The same underlying computational problem is solved across unrelated fields under different names: recursive B

深層学習軽量化・量子化予測テキスト

用途: 予測
難易度: Hard
コスト: High

Validating DBpedia Triple Sets for Natural Language Generation

We present a study of the quality of individual DBpedia triples from the perspective of Natural Language Gener

品質予測/異常検知自然言語処理生成テキスト

用途: 生成
難易度: Hard
コスト: Medium

We're Cooked! - Probing LLM Political Alignment Via Conflict-Framed Recipe Translation

Large language models (LLMs) are increasingly deployed for translation tasks, yet their implicit political pos

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Qwen-Audio-3.0-ASR Technical Report

In recent years, automatic speech recognition (ASR) has witnessed transformative advancements driven by three

自然言語処理大規模言語モデル分類テキスト音声

用途: 分類
難易度: Hard
コスト: High

コンピュータビジョンセグメンテーション画像テキスト音声

Where Should Language Sit in a Multimodal Model? Lessons from What Language Does to Human Perception and Cognition

Language models compute over tokens: language is their input, their output, and increasingly their internal re

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

MEMO: Multimodal Evidence Memory Organization for Long-Horizon LLM Agents

Long-running LLM agents rely on external memory to store and reuse information beyond a single context window,

深層学習軽量化・量子化画像テキストマルチモーダル

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

FramingQA: Does the Question Shape the Answer? Measuring the Compositional Framing Effect

We introduce FramingQA, a benchmark that measures the model sensitivity to question framing across law, medici

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Beyond Single-Negative Preference: Multi-Negative DPO for LLM-Centric Historical Entity Linking

Large language models (LLMs) have recently shown promise for historical entity linking, but preference optimiz

用途: 生成
難易度: Hard
コスト: High

CPUで試しやすいMI向き深層学習Transformer生成テキスト

Beyond Fluent Generation: A CPU Reliability Benchmark for MCP-Style Tool Calling in Sub-2B Small Language Models for Edge Deployment

Resource constrained single-board computers including Raspberry Pi, NVIDIA Jetson Nano, Arduino UNO Q, Orange

用途: 生成
難易度: Hard
コスト: Low

BlueprintAgent: Constraint-Triggered Targeted Revisits for Simulation-Ready Generation from Scanned Structural Blueprints

Converting in-service reinforced-concrete (RC) building blueprints into simulation-ready models---structured f

自然言語処理大規模言語モデル生成テキスト3D

用途: 生成
難易度: Hard
コスト: High

Content-Based Addressing for Long Context

Rotary position embedding (RoPE) uses each token's integer position to determine the rotation applied inside a

自然言語処理埋め込み・検索テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

LANTERN: Language Model Assessment on Noisy and Transformed Tasks for Understanding Error and Robustness Nuances

Robustness evaluation of large language models (LLMs) remains a critical challenge, particularly in assessing

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

SPARROW: Scalable Taxonomy Induction via Structure-Preserving Partitioning and Constraint-Guided Merging

Taxonomy induction aims to organize concept sets into coherent hierarchical structures. Recent LLM-based metho

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

CPUで試しやすい深層学習Transformerテキスト

RouteRelay: Event-Triggered Cross-Layer Route Reuse for Efficient Dynamic Sparse Attention

Dynamic sparse attention reduces long-context prefill cost by routing each query chunk to a small set of key c

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Probing the Structure and Dynamics of LLM Value Expression through Value Conflicts

Ethical evaluation of Large Language Models (LLMs) often characterizes model values as static and monolithic.

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

TAD: Token-Adaptive Contrastive Decoding with Confidence-Guided Gating for Hallucination Mitigation in Large Audio-Language Models

Large audio-language models (LALMs) can hallucinate audio objects, answering "yes" to absent sound events, thu

機械学習半教師あり学習QAテキスト音声

用途: QA
難易度: Hard
コスト: High

Separating Stream Stability from Long-Term Recall in Language Models

Methods for streaming language models are often discussed alongside long-context and memory systems, although

深層学習RNN / LSTM生成テキスト

用途: 生成
難易度: Hard
コスト: Low

CEDAR: Error-Bounded Residual Routing for Efficient Long-Context Attention

Post-hoc sparse attention accelerates long-context prefill by routing each query to a small set of token-level

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

KABURI-TTS: Phoneme-Keyed Activity-conditioned Bi-channel Utterance Rendering for Interaction

Realizing full-duplex spoken dialogue requires large amounts of two-channel, one-speaker-per-channel conversat

深層学習Transformer生成テキスト音声

用途: 生成
難易度: Hard
コスト: Medium

SIFTING: A Novel LLM-Based Framework for Structured and Transparent Information Extraction from Clinical Free-Text Reports, with Application to Tumor Staging in Lung Cancer

Background: Large language models (LLMs) show promise for extracting information from clinical free-text docum

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

In-Place Instruction Following in Diffusion Language Models

Diffusion Large Language Models (dLLMs) generate text via bidirectional iterative denoising, naturally support

用途: 生成
難易度: Hard
コスト: High

FreqBLiMP: Frequency-Controlled Minimal Pairs Reveal Robustness and Fragility of LLMs Under Lexical Rarity

Minimal-pair benchmarks such as BLiMP evaluate linguistic knowledge by testing whether language models (LMs) p

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Encoded Early, Used Late: Where Transformers Begin to Act on an Inferred Partner's Expertise

A transformer can make an attribute linearly decodable in its residual stream at a depth where that attribute

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Line-Coupled Language Model

Autoregressive language models generate one token per decoding step, limiting the useful output of each forwar

用途: 生成
難易度: Hard
コスト: High

The Illusion of Debiasing: Persona Steering Redistributes Rather Than Reduces Bias in LLMs

Prompt-based interventions: system prompts, personas, role instructions, reliably reshape what a language mode

用途: 生成
難易度: Hard
コスト: High

品質予測/異常検知深層学習軽量化・量子化テキスト強化学習

Revisiting Complete Reasoning Traces for Post-Training

Large language models (LLMs) are often post-trained on pre-collected reasoning trajectories to improve their r

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知深層学習軽量化・量子化検出生成QA

Where to Look and What to Use: Retrieve-Localize-Generate for Long-Term Conversational Memory Question Answering

Retrieval-augmented generation (RAG) enables large language models (LLMs) to answer questions by accessing ext

用途: 検出
難易度: Hard
コスト: High

A Hyperbolicity Atlas of Large Language Model Hidden States

LLM hidden states are ordinary vectors, but the distances among those vectors may still show hierarchical stru

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Aha-Flow Distillation: Flow Markers Matter in LLM Reasoning

We identify the Flow Moment, a reasoning pattern characterized by sustained, process-confirming verbalizations

MI向き深層学習軽量化・量子化テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデル検出生成テキスト

LatentMD: Benchmarking Markdown Boundary Failures in LLM-Generated Text

Large language models (LLMs) increasingly generate Markdown that is consumed by renderers, agents, code extrac

用途: 検出
難易度: Hard
コスト: High

Train Overcomplete, Deploy Compact: Scaling Recovery Capacity for Structured LLM Pruning

Large language models achieve strong performance across diverse tasks, but deployment remains costly because o

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

CantoneseLLM v2: Reasoning in a Low-Resource Language

Cantonese is widely spoken but remains low-resource in written data, with no large corpus of native Cantonese

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

MOLE: Detecting Insider Threats in AI Agents

Model misalignment, prompt injection, or operator misuse could lead AI agents operating frontier-lab accounts

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

TurEngMix: A Text Corpus and Benchmark for Turkish-English Code-Mixed Language Identification and Named Entity Recognition

Natural language processing systems underperform on code-mixed text, particularly for low-resource language pa

自然言語処理大規模言語モデル分類テキスト

用途: 分類
難易度: Hard
コスト: High

Dynamic-Programming-Guided Hierarchical BPE and Empirical Analysis of Vocabulary Pruning

Byte Pair Encoding (BPE) constructs vocabularies through greedy pair merging, but the resulting merge order do

深層学習軽量化・量子化セグメンテーションテキスト

用途: セグメンテーション
難易度: Hard
コスト: High

Towards Bridging the Gap Between Offline and Iterative Alignment via Preference Distillation

Direct preference optimization DPO is a promising offline approach for aligning large language models (LLMs) d

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Contextual Observer Grounding: Evaluating Situated Spatial Reasoning in Vision-Language Models

Reasoning over language instructions in embodied tasks such as robotics often requires understanding spatial r

深層学習Transformer検出テキスト3D

用途: 検出
難易度: Hard
コスト: High

Are Image Generators Zero-Shot Perceivers? A Rigorous Evaluation

Recent work, such as Vision Banana, shows that lightweight instruction tuning can enable an image generator to

深層学習軽量化・量子化生成セグメンテーション画像

用途: 生成
難易度: Hard
コスト: High

Scene Graph-Driven Haptic Feedback for Safety Enhancement in Robotic Ophthalmic Surgery via Physically Simulated iOCT

Robotic ophthalmic surgery offers high precision but introduces a "sensory gap" by decoupling the surgeon from

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

説明可能自然言語処理ファインチューニング検出画像テキスト

DroneGround: Open-Vocabulary Drone Payload Characterization Using Synthetic Data and Grounded Vision-Language Models

Automated drone surveillance has become increasingly important for public safety, critical infrastructure prot

用途: 検出
難易度: Hard
コスト: High

Zero-Shot 3D Plant Organ Segmentation with SAM3 and Semantic NeRFs

Accurate 3D plant organ segmentation is fundamental to automated phenotyping. Existing approaches rely on anno

自然言語処理RAGセグメンテーション画像テキスト

用途: セグメンテーション
難易度: Hard
コスト: High

MI向きコンピュータビジョンマルチモーダル生成画像テキスト

Better Call CineCrew: Consistent Ultra-Long Narrative-to-Film Generation

Long-form narrative-to-film generation requires shot-level controllability and cross-clip consistency in both

用途: 生成
難易度: Hard
コスト: High

BarkNet-Lite: A Lightweight Texture and Colour Network with the BarkBD Benchmark for Bark-Based Tree Species Recognition in Bangladesh

Tree species recognition supports forest inventory and biodiversity monitoring but still depends on scarce tax

深層学習軽量化・量子化分類画像テキスト

用途: 分類
難易度: Hard
コスト: Low

SphereSOD: Geometry-Structure Coupled Learning for 360 Salient Object Detection

360° salient object detection (SOD) aims to accurately segment salient regions across a full field of view. Ho

コンピュータビジョン物体検出検出テキスト

用途: 検出
難易度: Hard
コスト: Medium

深層学習Transformer分類異常検知テキスト

Mitigating Shortcut Learning: Texture-Penalized Prototype Networks

Standard Convolutional Neural Networks (CNNs) exhibit severe performance degradation due to a strong inductive

用途: 分類
難易度: Hard
コスト: Low

自然言語処理埋め込み・検索画像テキストマルチモーダル

CrACK: Adversarial Attacks on Cross-Model Consistency in Collaborative Vision Foundation Models

Training-free collaborative pipelines that integrate Vision Foundation Models such as CLIP, SAM, and DINO achi

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

説明可能コンピュータビジョン物体検出分類検出セグメンテーション

When Superpixels Fail on Documents: A Study of Segmentation for LIME Explanations

Post-hoc explanation methods are widely used to inspect image classifiers, but their reliability depends on de

用途: 分類
難易度: Hard
コスト: Low

Unified Vision-Centric Pedestrian Crossing Action Prediction via Adaptive Patch Projection and Proactive Spatial Rectification

Vision cues are available and informative for pedestrian action prediction, but obtaining stable target-centri

深層学習軽量化・量子化画像テキスト動画

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知深層学習Transformer分類テキスト動画

Federated Binary Gating with Server-Side Vision-Language Inference for Surveillance Anomaly Classification

Privacy-sensitive surveillance systems could benefit from large vision-language models (VLMs), but such models

用途: 分類
難易度: Hard
コスト: High

Social Intuition vs. Machine Reasoning: Anticipating Human-Robot Interaction from multiple modalities

Anticipating whether a person will interact from one's own perspective is a highly intuitive task for humans,

深層学習軽量化・量子化検出テキスト動画

用途: 検出
難易度: Hard
コスト: High

TRAIL: Trajectory-Aware Visual Place Recognition against Unordered Databases

Modern Visual Place Recognition (VPR) methods excel on standard benchmarks yet remain brittle in feature-poor

表形式向き深層学習軽量化・量子化分類検出画像

用途: 分類
難易度: Hard
コスト: High

DF26: We Cannot Tell Fake From Real Anymore

We introduce DF26, a novel benchmark for detecting AI-generated videos containing fully synthetic clips produc

コンピュータビジョンマルチモーダル検出生成画像

用途: 検出
難易度: Hard
コスト: High

D3ARC: Time-Critical Distributed Disaster Detection for Asynchronous Cooperative Multi-Robot Systems

Climate change is increasing the severity and unpredictability of natural disasters. In time-critical crises s

自然言語処理RAG検出テキスト

用途: 検出
難易度: Hard
コスト: Low

MV-STRIDE: Enabling MLLMs to Master Multi-View Spatial Reasoning via Hierarchical Capability Modeling

Despite the rapid progress of Multimodal Large Language Models (MLLMs) in 2D vision-language tasks, robust mul

自然言語処理大規模言語モデル生成テキスト3D

用途: 生成
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理プロンプトエンジニアリング検出異常検知画像

Proximity-CLIP: Text-Guided Semantic Proximity Learning for Zero-Shot Anomaly Detection

Vision-language models offer a promising approach for zero-shot anomaly detection (ZSAD). However, due to obje

用途: 検出
難易度: Hard
コスト: High

The Price of Consistency: Exploiting Visual Anchors for Multimodal Jailbreaking in Video Generation

The rapid evolution of video generation has shifted the paradigm from pure text-driven to multi-conditional co

コンピュータビジョン動画認識生成画像テキスト

用途: 生成
難易度: Hard
コスト: High

Unsupervised Domain Adaptation for Symbol Spotting in Historical Encrypted Manuscripts

The decipherment of historical encrypted manuscripts poses a fundamental challenge in Digital Humanities: befo

深層学習Transformerテキスト教師なし

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

自然言語処理埋め込み・検索テキスト動画マルチモーダル

Ambient @ EgoLongQA 2026: Distilling Long-Video perception into a Sub-2B Model

We describe our entry to the EgoLongQA track of the Wearable-AI Challenge in ECCV 2026, which placed first in

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知深層学習軽量化・量子化テキストマルチモーダル強化学習

Stable-MM-R1: Anchoring Multimodal Reasoning Dynamics via Entropy-Guided Stratification

While Reinforcement Learning (RL) effectively incentivizes reasoning in Large Language Models, current pipelin

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知深層学習Transformer生成画像テキスト

Flow3D-OPD: Multi-Teacher On-Policy Distillation for 3D Geometry Generation with Flow-Matching Diffusion Transformer

Recent image-to-3D generation models built on flow-matching diffusion Transformers (DiT) can produce high-fide

用途: 生成
難易度: Hard
コスト: High

コンピュータビジョンマルチモーダル画像テキスト動画

NutriBench-Kitchen: Benchmarking Embodied AI for Nutrition Management

An embodied kitchen assistant must do more than recognize food in isolated frames. It must track ingredient st

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

説明可能品質予測/異常検知深層学習軽量化・量子化分類生成画像

Discovering Natural Transformation Vulnerabilities in Black-Box Vision Models

Natural adversarial examples (NAEs) reveal that vision models can fail under realistic semantic changes beyond

用途: 分類
難易度: Hard
コスト: High

Ambient @ EgoProactive 2026 : Proactive Egocentric Assistance with Visually Grounded Supervision

We present our submission to the EgoProactive track of the ECCV 2026 Wearable AI Challenge, which ranked first

コンピュータビジョンマルチモーダル分類生成画像

用途: 分類
難易度: Hard
コスト: High

CASCADE: A Spatio-Temporal-Causal Reasoning Representation and Dataset for Driving

Reasoning is a promising route to the generalization that autonomous driving requires in the long tail, as it

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理ファインチューニング生成画像テキスト

Single Image to Textured 3D Object Generation in Frequency Domain: From Theory to Pipeline

Single-view 3D reconstruction, also known as image-to-3D, is a persistently challenging task due to the extrem

用途: 生成
難易度: Hard
コスト: High

表形式向き自然言語処理大規模言語モデル分類検出テキスト

SupGRPO: Enhancing GRPO with Matching-based Online SFT for Text Spotting

Text spotting requires both accurate text recognition and precise spatial localization. Current specialised sp

用途: 分類
難易度: Hard
コスト: High

センサ/時系列深層学習Transformer埋め込みテキスト3D

One for All: Generalist Foundation Model for Cross-Sensor Skeleton Representation Learning

For learning generalizable motion representations from large-scale unlabeled data, Self-supervised learning (S

用途: 埋め込み
難易度: Hard
コスト: High

品質予測/異常検知コンピュータビジョンセグメンテーション画像テキスト

Continuous Token-Level Spatio-Temporal Context Modeling for Visual Object Tracking

Spatio-temporal context has become increasingly crucial for visual tracking. However, most existing approaches

用途: 技術検証・論文読解補助
難易度: Easy
コスト: Medium

SpatialBlock: Enhancing Spatial Intelligence in LVLMs via Synthetic Block-Stacking Problem

Large Vision-Language Models (LVLMs) have achieved strong performance on diverse visual tasks, yet their abili

自然言語処理RAGQA画像テキスト

用途: QA
難易度: Hard
コスト: High

品質予測/異常検知コンピュータビジョンマルチモーダル生成QA画像

FreqDoor: A Hidden Trojan in the Frequency Domain for Backdoor Attacks on Vision-Language Models

Vision-language models (VLMs) have recently shown excellent progress in open-ended image-to-text generation. H

用途: 生成
難易度: Hard
コスト: High

LoGAN: Multilingual Font Localization with Generative Agents

Localizing a font into new languages is a highly intricate task requiring precise design adaptation of glyphs,

少数データ向き自然言語処理RAG検出生成画像

用途: 検出
難易度: Hard
コスト: High

品質予測/異常検知深層学習Transformer画像テキスト動画

Fine-Grained Visual Preprocessing and Dual-Stream Temporal Modeling for Multimodal Sentiment Analysis on Social Media

Multimodal sentiment analysis often remains text-dominant due to raw-video noise and insufficient temporal mod

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

NeuCME: Toward Dynamic Multimodal Continual Learning via Neural Combinatorics of Multiple Experts

Multimodal continual learning has recently shown great potential for developing agents with human-like intelli

深層学習軽量化・量子化テキストマルチモーダル

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

自然言語処理大規模言語モデル検出セグメンテーション画像

CGSM: Concept-Guided Segmentation Model for Precise Pulmonary Lesion Delineation

Accurate segmentation of pulmonary lesions is essential for effective clinical diagnosis and treatment strateg

用途: 検出
難易度: Hard
コスト: High

AV-SafetyBench: A Safety Benchmark for Text-to-Audio-Video Generation

Recent text-to-audio-video (T2AV) models jointly generate video, speech, sound effects, and ambience from a si

生成AI動画生成生成画像テキスト

用途: 生成
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデル検出生成異常検知

AnomalyCraft-700K: Component-Level Controllable and Verifiable Synthetic Anomalies for Fine-Grained Video Anomaly Understanding

Progress in video anomaly understanding (VAU) has long been limited by inherent deficiencies of real-world ano

用途: 検出
難易度: Hard
コスト: High

少数データ向き自然言語処理プロンプトエンジニアリング画像テキストマルチモーダル

Re-calibrated Contrastive Loss for Transformation-Aware Prompt Conditioning in Vision-Language Models

Ensuring effective transfer learning for vision-language models without compromising their generalization perf

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

センサ/時系列品質予測/異常検知深層学習軽量化・量子化画像テキストマルチモーダル

M3-Tele: A Unified Multimodal Teleoperational Framework for Compliant Whole-Body Mobile Manipulation

Executing contact-rich tasks efficiently requires the seamless integration of whole-body coordination and phys

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

少数データ向き自然言語処理RAG生成テキストマルチモーダル

ICI-VLA: In-Context Imitation with Spatiotemporally Aligned Demonstrations for Vision-Language-Action Models

Vision-Language-Action (VLA) policies are commonly adapted to new manipulation settings through additional gra

用途: 生成
難易度: Hard
コスト: High

センサ/時系列自然言語処理RAG画像テキストマルチモーダル

Zero-Shot Sim-to-Real Contact-Rich Assembly via Proprioception-Anchored Cross-Modal Pretraining

Contact-rich assembly remains challenging because it requires submillimeter spatial accuracy and reliable inte

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Functional-SLAM: Interaction-Aware Mapping with Online Functional Scene Graphs

Existing SLAM systems lack modeling of the functional relations required for fine-grained robotic interaction.

深層学習軽量化・量子化画像テキスト3D

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Measuring Language Transfer in Robot Policies: Adding Greek to a Cosmos3 Vision-Language-Action Policy

Robot foundation models are trained and evaluated predominantly in English, and robot demonstration corpora do

コンピュータビジョンマルチモーダル検出テキスト

用途: 検出
難易度: Hard
コスト: High

PV-WM: A Heterogeneous Micro-Macro World Model for Articulated Pedestrian-Vehicle Co-Rollout

Local pedestrian-vehicle forecasting spans heterogeneous physical scales: pedestrians combine root locomotion

深層学習RNN / LSTM予測テキスト

用途: 予測
難易度: Hard
コスト: Low

センサ/時系列自然言語処理大規模言語モデルテキスト

How Long Until Your Robot Ignores You? A Safety Benchmark for LLM Orchestrators in Human-Humanoid Collaboration

Large Language Models (LLMs) are increasingly employed to orchestrate robot behavior through natural-language

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Generalizable 6D Pose Estimation of Textureless Objects with Planar-based Gaussian Splatting

Estimating the 6D pose of textureless objects without prior CAD models remains a critical challenge due to the

自然言語処理RAG画像テキスト3D

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Phase-and-First-Arrival VLM Feedback for Sparse-Reward Reinforcement Learning in Surgical Manipulation

Sparse outcome feedback limits what robots can learn from unsuccessful attempts at complex manipulation. Faile

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

MAC-I$^2$: Learned Metrics-Aware Covariance for Robust Visual-Inertial Fusion in Initialization and Calibration

Visual-Inertial (VI) fusion is fundamental to accurate and robust state estimation, where camera and IMU measu

深層学習Transformer画像テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

説明可能自然言語処理大規模言語モデルテキスト動画強化学習

From LLM-Generated Specifications to Learned Quadruped Locomotion

Quadruped robot locomotion policies are often trained using reinforcement learning, which in turn relies heavi

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

RoboDreamer: Anticipatory Humanoid Locomotion with Predictive State-Space Models

Humanoid locomotion requires control policies that remain stable under imperfect sensing while exploiting temp

センサ/時系列深層学習軽量化・量子化テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

MEMOBench: A Process Level Memory Benchmark for Robotic Manipulation

Robotic manipulation often requires acting on information that is no longer visible, yet Vision-Language-Actio

自然言語処理RAGテキストマルチモーダル

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

センサ/時系列深層学習RNN / LSTMテキストマルチモーダル

WM-Craftnet: World Synesthesia Model for Generalizable and Robust Dexterous In-Hand Manipulation

Generalizable and robust dexterous in-hand manipulation requires a policy to infer object pose, geometry, cont

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

The Internal Anatomy of Strategic Choice in Large Language Models

Large language models act as strategic agents and models of human choice, yet choosing like a strategic agent

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

表形式向きCPUで試しやすい自然言語処理RAG分類検出回帰

Protocol effects on feature-based hardware-Trojan detection across Trust-Hub families

Trust-Hub reuses host circuits: several files differ mainly in the inserted Trojan. When gates from sibling va

用途: 分類
難易度: Hard
コスト: High

Beyond Worst-Case Coreset Bounds for $k$-Clustering via Determinantal Sampling

Massive datasets in modern machine learning have made data reduction a central challenge, particularly for clu

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Formation of structural attractors in neuromorphic systems

This paper examines the theory of Invariant Structural Learning (ISL), which proposes a non-optimization appro

生成AIGAN分類画像テキスト

用途: 分類
難易度: Hard
コスト: High

Linear Algebra Foundations of Efficient Attention: A Phase Reversal in Rank Collapse Under SVD Compression

Linear algebra provides the framework of concepts (matrix rank, singular value decomposition (SVD), and eigend

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

You Are What You Read: Misalignment via In-Context Persona Induction

Broad misalignment has been produced by finetuning on narrow data, harmful or benign, and in context only by d

自然言語処理プロンプトエンジニアリングテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

品質予測/異常検知自然言語処理大規模言語モデル分類画像テキスト

NormViz: A Benchmark and Framework for Grounding Multimodal Reasoning in Global Cultures

AI systems are used worldwide, but they struggle to serve the needs of culturally diverse populations. Prior w

用途: 分類
難易度: Hard
コスト: High

Typed Federated Artifacts for the Agentic Web:Sharing Tool-Routing Knowledge Across Frozen,Heterogeneous LLM Agents

An open, networked web will allow agents to run frozen models from multiple vendors, keep their history privat

自然言語処理大規模言語モデル分類テキスト

用途: 分類
難易度: Hard
コスト: High

Measuring GEO Visibility: Prompt Corpora Define the Answer Market

GEO (generative engine optimization) visibility scores aggregate source appearances, citations, or brand menti

深層学習Transformer生成検索テキスト

用途: 生成
難易度: Hard
コスト: Low

少数データ向き条件最適化自然言語処理ファインチューニング分類テキスト

Train Smarter, Not Harder: Switching Signal-Guided Training in Active Learning

Training strategy, namely whether to retrain from scratch or fine-tune from the previous checkpoint, is an ove

用途: 分類
難易度: Hard
コスト: High

AuthBench: A Large-Scale Multilingual Benchmark for Authorship Representation across Genres and Lengths

Authorship signals matter in settings where writing style carries identity: digital forensics, plagiarism anal

自然言語処理RAG検出テキスト

用途: 検出
難易度: Easy
コスト: Low

A Novel Semantic Manifold Alignment Attack against Embedding-to-Embedding Obfuscation in Privacy-Preserving LLMs

With the widespread applications of large language models (LLMs), privacy-preserving inference has become incr

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

深層学習RNN / LSTM画像テキストマルチモーダル

Reason Through the Latent! Making Latent Visual Reasoning Necessary

Latent visual reasoning aims to perform multimodal reasoning through hidden-state computation rather than expl

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

自然言語処理プロンプトエンジニアリング画像テキスト

Counterfactual Tests for Measuring Chain-of-Thought Faithfulness in Visual Language Models

Chain-of-thought (CoT) may often look plausible, yet it may not faithfully reflect the model's decision-making

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

品質予測/異常検知自然言語処理RAG生成画像テキスト

DianShi-RxnDB: A Large-Scale, Fine-Grained Organic Reaction Data Platform Built via a Fully Automated Pipeline for Researchers and AI Agents

High-quality structured organic reaction data are essential for developing artificial intelligence for chemist

用途: 生成
難易度: Hard
コスト: Low

PARSER: Read in Parallel, Reason in Depth for Long-Context LLM Agents

Sequential memory agents process long documents by reading chunks one after another while maintaining a compac

深層学習Transformerテキスト強化学習

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

A Grapheme-Aware Indic Tokenizer for Tamil: Large-Scale Training and Intrinsic Evaluation

Tokenization forms the foundation of modern Natural Language Processing (NLP) systems by transforming raw text

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Data Efficient Sample Selection for In-Context Learning

The In-context learning (ICL) paradigm aids large language models (LLMs) to adapt to new tasks without need fo

説明可能深層学習軽量化・量子化テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

EviMap: Evidence-Grounded Hierarchical Topic Maps for Exploring Unlabeled Corpora

Research teams and organizations often explore unfamiliar free-text collections, from survey comments and revi

自然言語処理大規模言語モデルテキスト動画

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

表形式向き自然言語処理大規模言語モデルテキストマルチモーダル

ECOKV: Geometry-Aware KV Cache Eviction via Complementary Diversity Metrics

Although multimodal Large Language Models (MLLMs) excel in diverse tasks, their scalability remains limited by

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

SerenAI: State-transition system inspired by text-based world AI models

Although professional workflows leverage large language models widely, the interpretation for auditing unconst

深層学習軽量化・量子化生成テキスト

用途: 生成
難易度: Hard
コスト: High

Mind the Gap: Exposing LLM Translation Blind Spots Using the AlphaMWE Multilingual Parallel Corpus

LLMs' performance on machine translation (MT) tasks is often dependent on the data availability in the specifi

深層学習Transformer翻訳テキスト

用途: 翻訳
難易度: Hard
コスト: High

SAGE: A Hierarchical Framework for Evaluating Interpretive Literary Quality in Narratives

Assessing the literary quality of narratives requires evaluating interpretive dimensions (cultural representat

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Discovering Translation-Worthy Languages with E-Values

Choosing when to translate multilingual documents is a central routing problem in text classification: transla

機械学習教師あり学習分類テキスト

用途: 分類
難易度: Hard
コスト: Low

A Translational Note on AI Safety Evaluation

Recent studies report that automated red-teaming finds more vulnerabilities, at lower cost, than human red-tea

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

LLMs Mirror Country-Specific Gender Patterns If Asked, but Skew Male When Generating Media in Local Languages

Large language models (LLMs) are increasingly used to generate media, but whether their content perpetuates ge

用途: 生成
難易度: Hard
コスト: High

ProcArena: A Multi-Scenario Benchmark for LLMs on Direct and Interactive PL/SQL Development from Natural Language

Large language models (LLMs) have shown strong potential for translating natural-language (NL) requirements in

用途: 生成
難易度: Hard
コスト: High

One Step, One Lead: Mitigating Higher-Order Interference in Multi-Domain Reinforcement Learning via Cross-Step Control

Reinforcement learning (RL) across multiple domains can broaden the reasoning capabilities of large language m

自然言語処理大規模言語モデルテキスト強化学習

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理大規模言語モデル生成画像テキスト

InsightChain: Optimized Chain-of-Insight Analytics for LLM-driven Data Visualization

Large language models (LLMs) are increasingly used for automated data visualization, yet existing approaches o

用途: 生成
難易度: Hard
コスト: High

品質予測/異常検知自然言語処理RAG分類画像テキスト

Visual Search Augmented Chain-of-Thought Reasoning for Attribute Value Extraction from Product Videos

Existing approaches to visual attribute value extraction (AVE) primarily rely on static product images, failin

用途: 分類
難易度: Hard
コスト: High

Building Trustworthy Graph-Agentic RAG for Social Good: Architectures, Failure Propagation, and Assurance by Construction

Graph-agentic retrieval-augmented generation combines structured evidence with adaptive controllers that can p

品質予測/異常検知自然言語処理RAG生成テキスト

用途: 生成
難易度: Hard
コスト: Low

Cross-Lingual Representation Alignment by Token-Level Optimal Transport in a Language-Agnostic Space

Cross-lingual alignment (CLA) aims to align the representations of large language models (LLMs) across languag

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

A Ticket from Marginals to Joints: Coupled-Noise Distillation for One-Step Block Generation in Diffusion Language Models

Autoregressive language models commit one token per forward pass; diffusion language models commit a block of

用途: 生成
難易度: Hard
コスト: High

ContextFlow: In-Context Flow Matching for Robot Manipulation

Although highly effective in vision and language domains, applying in-context learning to robotics remains cha

自然言語処理RAG生成画像テキスト

用途: 生成
難易度: Hard
コスト: High

Physico-Geospatial Grounded Scene Interpretation for Mobile Robotics

Recent advancements in deep learning allow robotic agents to interact with dynamic and unstructured environmen

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Unifying Physics-Based Humanoid Interaction with a Context-Conditioned Interaction Prior

Developing unified physics-based humanoid controllers that can navigate complex 3D scenes and manipulate objec

コンピュータビジョン3D・点群テキスト3D

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

\textbf{PLATO}: \emph{Preintegration Learning from Accurate Trajectory Observations} for Neural Inertial Odometry

Neural inertial odometry has demonstrated strong potential for motion estimation in challenging environments,

深層学習軽量化・量子化画像テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

Can People Distinguish Human and AI Agency in Humanoid Teleoperation? A Preliminary Study of Agency Perception

Can people distinguish between human and AI agency in humanoid teleoperation? To explore this question, we dev

コンピュータビジョン動画認識テキスト音声

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Rethinking Safety for Generalist Robots

Generalist robots promise to transform our society: the same system that prepares a meal or folds laundry migh

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

arxivPaper only2026-09-05

Online Learning with LLM Experts from Limited Feedback

We study adaptive routing of prompts to large language model (LLM) experts to maximize response quality in an

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-09-05

Minimizing the Effect of Sleep Deprivation in the Forward-Forward Algorithm

This paper addresses the challenge posed by sleep deprivation in the Forward-Forward algorithm, where separati

深層学習テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

arxivPaper only2026-09-05

RefGuard: Identity-Aware Language-Guided Robot Manipulation via Joint Target-Anchor-Frame Grounding

Vision-language-action (VLA) models have substantially advanced language-guided robot manipulation, yet reliab

自然言語処理大規模言語モデルテキストマルチモーダル

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

コンピュータビジョンセグメンテーション画像テキスト教師あり

An Analysis of Self-supervised Pre-training with Dependent Samples

Self-supervised learning relies on so-called data augmentations $φ(x)$ of unlabeled datapoints $x$ --- for exa

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

コンピュータビジョンセグメンテーション生成テキスト

What Makes a Redundant Representation Remember? Lineage Isolation, Not Masking

Memory-based evolutionary algorithms for dynamic optimization often carry a redundant second copy of the genot

用途: 生成
難易度: Hard
コスト: Medium

Large Language Models with At Most One Spike per Neuron

Leveraging their inherent sparse event-driven computation, spiking neural networks (SNNs) offer a promising pa

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

自然言語処理大規模言語モデル画像テキストマルチモーダル

Time-Aware Assistive Navigation

Can interactive vision-and-language agents learn not just what to say but also \textbf{\textit{when}} to say i

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Same Trajectory, Contradictory Rewards (ROBORMBENCH): Paraphrase Fragility in Vision Language Reward Models

Vision-language models are increasingly used as reward functions for robotic learning, but this role requires

自然言語処理RAGテキストマルチモーダル

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

センサ/時系列深層学習Transformer回帰予測テキスト

Adaptive Cost-Sensitive Machine Learning for Autonomous Robot Navigation Failure Prediction: When Not All Errors Are Equal

Autonomous robot navigation failures differ not only in categorical severity but also in the physical context

用途: 回帰
難易度: Hard
コスト: High

CrossDepth: Geometry-Constrained Attention for Generalizable Multi-View Surround Depth Estimation

Reliable 3D understanding of the surrounding environment is a core requirement for autonomous driving. Multi-v

自然言語処理RAG画像テキスト3D

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Towards Neuro-Symbolic Procedural Reasoning for Long-Horizon Vision-Language-Action Manipulation

Vision-language-action (VLA) models can execute short manipulation skills, but remain brittle in long-horizon

自然言語処理ファインチューニング画像テキスト動画

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Development of a Humanoid Robot Prototype for Multimodal Human-Robot Interaction

Human-robot interaction (HRI) enables intuitive and intelligent collaboration between humans and robots in rea

深層学習RNN / LSTM分類検出テキスト

用途: 分類
難易度: Hard
コスト: High

arxivGitHubあり2026-09-04

RoboSPA: Can VLA Models Go Beyond Simple Scenes and Short-Horizon Tasks?

Vision-Language-Action (VLA) models have shown promising progress in language-conditioned robotic manipulation

コンピュータビジョンマルチモーダルテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

深層学習Transformerテキストマルチモーダル

LIBERO-RECOVER: Beyond Task Success Towards Failure Recovery in Robotic Manipulation Models

Vision-Language-Action (VLA) or World Action (WAM) models have recently demonstrated remarkable performance in

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

A Schema Bounded Language Model for Refining Robot Policies Without Destabilizing Local Learning

This paper addresses navigation by composite heterogeneous robots in a decentralized system when policy reason

用途: 生成
難易度: Hard
コスト: High

One Diffusion Model, Two Roles: Guided Trajectory Planning and Safety-Critical Scenario Generation in Closed-Loop Simulation

Diffusion probabilistic models can capture the multi-modal, interaction-rich distribution of joint future traj

用途: 生成
難易度: Hard
コスト: High

Reasoning Without Inference Cost: Latent Semantic Scaffolding for Robot VLA Policies

Vision-language-action (VLA) models are trained by imitation and capture what action to take but not why; addi

自然言語処理埋め込み・検索テキストマルチモーダル

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Continuous Cognitive Coverage for Autonomous Robots via Event-Dependent Cognitive Treatment and Learning

Autonomous robots continuously encounter objects, changes, and situations, and every event admitted into cogni

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

Open-Set 3D Scene Graphs for Field Robotics: An Outdoor Case Study

Three-dimensional scene graphs (3DSGs) have emerged as a promising approach for building geometrically grounde

深層学習軽量化・量子化テキスト3Dマルチモーダル

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

A Constraint-Aware Generative Framework for Synthetic Origin-Destination Demand in Logistics Networks

Large-scale logistics networks require synthetic data generation capabilities to support scenario-based planni

深層学習グラフニューラルネット生成テキスト

用途: 生成
難易度: Hard
コスト: High

ALRA: Adaptive Local Relational Alignment for Logit-Based Pre-training Distillation of Autoregressive Language Models

Logit-based knowledge distillation for autoregressive language models usually aligns teacher and student next-

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

How Perturbations Propagate: A Multi-Level Analysis of Robustness in Large Language Models

Language models encounter typos, corrupted text, altered words, and disrupted token order, yet robustness is u

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

説明可能MI向き品質予測/異常検知コンピュータビジョン物体検出検出画像テキスト

arxivGitHubあり2026-09-03

Understanding Autonomous Driving Datasets by Describing Differences between Image Subsets in Natural Language

Understanding the composition of large-scale autonomous driving datasets is essential for safety, robustness,

用途: 検出
難易度: Hard
コスト: High

品質予測/異常検知コンピュータビジョン3D・点群分類画像テキスト

Where Appearance Fails, Geometry Recognizes: A CAD-Free 3D Shape Prior That Complements Vision Foundation Models

Recognizing specific objects onboarded without a labeled training set recurs across manufacturing and service

用途: 分類
難易度: Hard
コスト: High

Adaptive Vision-Language Grasping via Composable Foundation Priors and Generalizable Grasp Synthesis

This paper proposes AdaRoboVLG, a task-adaptive Vision-Language-Grasp (VLG) framework that supports generaliza

深層学習軽量化・量子化生成テキスト動画

用途: 生成
難易度: Hard
コスト: High

Continuous Actions from Discrete Minds: Latent-Aligned Planning for End-to-End Autonomous Driving

Bridging the gap between the discrete reasoning of Vision-Language Models and the continuous, physics-constrai

深層学習Transformer生成画像テキスト

用途: 生成
難易度: Hard
コスト: High

Toward Unified Robot Learning: Bridging Representation, Vision-Language-Action, and World Models

For robots to operate reliably in real-world environments, they need to perceive their surroundings, act, and

自然言語処理RAG異常検知埋め込みテキスト

用途: 異常検知
難易度: Hard
コスト: High

深層学習軽量化・量子化テキストマルチモーダル強化学習

Revisiting Topological Graphs for Macro Action based Closed-loop Reinforcement Learning of Vision Language Navigation in Continuous Environment

Vision-Language Navigation in Continuous Environments (VLN-CE) requires an agent to follow natural language in

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivGitHubあり2026-09-03

Abstraction Agent

Information abstraction, which groups strategically similar private states into a tractable number of buckets,

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Unifying Conformal Language Tasks with In-Context Ensembles

Many NLP tasks, such as summarization and extractive question answering, reduce to retrieving relevant content

自然言語処理大規模言語モデル要約QAテキスト

用途: 要約
難易度: Hard
コスト: High

Causal Foundation Models

Causal inference is the practice of estimating the effect of a treatment or intervention from data. It traditi

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Self-Supervised Pretraining of Molecular Graph Encoders with LeJEPA

Self-supervised pretraining has transformed language and vision, but its value for molecular graph neural netw

深層学習Transformerテキスト自己教師

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

The Geometry of Ignorance: LLMs Know When to Temper Bayesian Priors

What does a language model predict when it has few clues? The answer lurks in its unembedding geometry: a sing

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

自然言語処理プロンプトエンジニアリングテキスト3D

arxivGitHubあり2026-09-02

H3DNAS: Hardware-Aware ONNX-Native 3D Point Cloud Model Compression

Deploying 3D point cloud models on edge hardware such as the NVIDIA Jetson Orin Nano is severely constrained b

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Semantics-Guided Automatic Tensorization for Multiobjective Evolutionary Algorithms: A Multi-Agent Framework

Multiobjective evolutionary algorithms (MOEAs) naturally expose population-level parallelism, but many mature

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

LLM-Driven Joint Evolution of Coupled Heuristics Components for Routing Optimization

Heuristic design for combinatorial optimization remains heavily reliant on expert knowledge, while existing la

用途: 生成
難易度: Hard
コスト: High

Fair Stable Matching: A Nash Social Welfare Approach

While traditional stable matching algorithms, such as the Gale-Shapley algorithm, prioritize stability, they m

機械学習特徴量エンジニアリングテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

表形式向き深層学習Transformer分類テキスト表形式

Measuring consistency via ensemble margin and local prediction variability: Auditing decision systems in the presence of predictive multiplicity

The Rashomon effect is a machine learning phenomenon where equally accurate models produce different predictio

用途: 分類
難易度: Hard
コスト: High

One-Layer Transformer Provably Learns Multiclass One-Nearest Neighbor in Context

We extend recent work establishing an equivalence between one-layer transformers and nearest-neighbor classifi

深層学習Transformer分類テキスト

用途: 分類
難易度: Hard
コスト: Low

深層学習軽量化・量子化セグメンテーションテキスト音声

arxivGitHubあり2026-09-01

Artificial Rosetta Stone: Constrained Maximum A Posteriori (MAP) Reconstruction of Symbolic Raga Sequences via Order-k Markov Models

Reconstructing a damaged musical fragment is an inverse problem: the observed sequence contains partial inform

用途: セグメンテーション
難易度: Easy
コスト: Low

コンピュータビジョンセグメンテーション生成テキスト

From Truncation to Commitment: Persistent Context in Uniform Discrete Diffusion

Uniform-state discrete diffusion models update all tokens in parallel while keeping every position revisable.

用途: 生成
難易度: Hard
コスト: High

自然言語処理大規模言語モデルテキスト音声マルチモーダル

Embedded Conditional Independence Tests for Large Language Model Generated Text with an Application to German Parliament Speeches

Conditional independence tests (CITs) test for conditional dependence between two random objects $X$ and $Y$ g

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

When Metropolis and Hastings Meet Bradley and Terry: Exact MCMC From Preference Voting

Sampling from distributions conditioned on desired semantic properties is an emerging challenge in modern gene

自然言語処理大規模言語モデル生成画像テキスト

用途: 生成
難易度: Hard
コスト: High

Reinforcement learning to choose optimizers

No single optimization method is uniformly best for all problems, and the most suitable optimizer choice can c

深層学習RNN / LSTMテキスト強化学習

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Web Price Extraction: State of the Art and an Adaptive Browserless Implementation

Price extraction from websites is a key task for market monitoring, price comparison, and business analytics i

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Denoising Diffusion Generative Models Secretly Calculate Attentions

Denoising diffusion models are the dominant architecture for image generation, whereas most natural language g

深層学習Transformer生成画像テキスト

用途: 生成
難易度: Hard
コスト: High

Self-Reports Are Not Verification: Environment-Grounded Auditing of LLM Operators in Evolutionary Search

Language model agents increasingly propose actions, observe external feedback, and explain their own behavior.

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Construction of a DFA for Computing Grundy Numbers in the Successful Derivation Games on Right-Linear Grammars

Inoue et al. have introduced the successful derivation game (SDG) on context-free grammars (CFGs), which is a

深層学習RNN / LSTMテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

When Can We Work in Embedding Space? What Text Embeddings Preserve

When do text embeddings work as inputs to empirical analysis? Their use rests on an assumption: that we can tr

説明可能自然言語処理大規模言語モデル生成テキスト

用途: 生成
難易度: Hard
コスト: High

強化学習方策勾配 (PPO / A3C)分類テキスト

Fairness in multi-class multi-group classification problems via contextial coherent risk measures

We propose a new design of fair classifiers for multi-class classification problems in the presence of vector-

用途: 分類
難易度: Hard
コスト: Low

Where Should Experience Live? Hierarchical Hebbian Memory for Continual Vision Transformers

Vision Transformers provide strong visual representations but typically rely on slowly updated parameters, lim

深層学習Transformer分類画像テキスト

用途: 分類
難易度: Hard
コスト: Low

Event-Driven Language Models with Sparse Neural Activity for Neuromorphic Hardware

Inference with transformer-based large language models (LLMs) is often limited by the memory-bound KV cache an

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Constant Individual Regret in General Games

Uncoupled no-regret dynamics provide a decentralized route to equilibrium, but prior guarantees for individual

深層学習正規化・最適化手法テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

表形式向きコンピュータビジョンセグメンテーションテキスト表形式強化学習

Test-time Reinforcement Learning in Imperfect Information Games

Test-time reasoning has significantly improved performance in domains ranging from games to language models. H

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

LangBP: Language-Guided Reasoning and Acting for Joint Bidding and Pricing

Auto-bidding is a long-horizon sequential decision problem for maximizing conversion value under budget and ke

用途: 生成
難易度: Hard
コスト: High

Mechanism Design for Facility Location Games Under a Prelocated Facility

We study the problem of locating a new homogeneous facility under a prelocated facility. Here, a set of $n$ ag

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Learning Representations through Token Prediction: Geometry, Approximation, and Downstream Guarantees

Token prediction is a central pre-training objective for modern language models. Despite its empirical success

深層学習RNN / LSTM分類生成テキスト

用途: 分類
難易度: Hard
コスト: High

Which LLM for Which Work? Budgeted Model Allocation under Uncertain Evaluation

A company with a fixed artificial intelligence (AI) budget must decide which large language model (LLM) handle

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Evolutionary Soups: Evolving Mixture-of-Experts for Multi-Objective LLM Alignment

Large language models are increasingly required to generate responses that satisfy multiple competing objectiv

深層学習軽量化・量子化生成テキスト

用途: 生成
難易度: Hard
コスト: High

説明可能品質予測/異常検知深層学習軽量化・量子化テキスト

PruneShift: A Framework for Evaluating Decision Reliability in Structured Pruning

Structured pruning uses surrogate objectives because direct task evaluation over every feasible mask is too ex

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

arxivPaper only2026-08-29

Deciding When to Decide: Testing Operational Suboptimality Under Distributional Shift

Deployed decisions are often optimized once and retained because updates impose operational, regulatory, or sw

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

arxivPaper only2026-08-29

Content Exploration Beyond the Feed: Creator Supply and the Shared Corpus

Industrial recommenders give new content initial views through budgeted exploration, then use early performanc

自然言語処理ファインチューニングテキスト動画

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-08-29

Jigsaw-CRL: Recovering Global Latent Causal Order from Fragmented Multi-Client Interventions

Causal representation learning (CRL) aims to recover latent causal variables and their structural relations fr

自然言語処理RAG埋め込みテキスト

用途: 埋め込み
難易度: Hard
コスト: Low

Learning a Size-Weight Frontier for Synthetic-Augmented Inference

Synthetic data can improve statistical inference when real data are scarce, but naively treating synthetic sam

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知深層学習軽量化・量子化検出異常検知テキスト

Localizing Global Discrepancies: Marginal Contributions and Contextual Anomaly Detection

Global goodness-of-fit and discrepancy statistics can establish that a sample departs from a reference distrib

用途: 検出
難易度: Hard
コスト: Medium

Fine-Tuning Autobidders with Group Relative Policy Optimization

Automated bidding (autobidding) is a core component of modern online advertising systems. Within this componen

自然言語処理大規模言語モデル生成テキスト強化学習

用途: 生成
難易度: Hard
コスト: High

AI Alignment through a Game-theoretic Lens: A Survey

As large language models and increasingly capable AI agents are deployed in high-risk settings, aligning them

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Fast Weight Attention for Continual Learning

Recurrent fast-weight memories and selective state-space models compress an expanding context into a fixed-siz

深層学習RNN / LSTM回帰テキスト

用途: 回帰
難易度: Hard
コスト: Medium

The role of parameter Jacobians in the stability of network outputs

In the framework of network dynamics, learning models, and neural tangent kernels (NTK), we show that the corr

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

説明可能センサ/時系列コンピュータビジョンセグメンテーションテキスト音声

Recovering Expert Critic-Sourced Network Adjacency between Musical Artists from Acoustic Distributions: A Construct-Validity Approach

Music recommendation relies primarily on two signals: user-item interactions, which fail in the cold-start reg

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Representation Measurements Under Function-Preserving Reparameterizations

Hidden coordinates are not uniquely determined by a language model's input--output function, so representation

MLOpsモデルデプロイテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

Bug Localization from Bug Reports: A Multi-Objective Approach

Bug localization is a labor-intensive task, particularly in large software systems. When abnormal behavior occ

数学・理論最適化検出テキスト

用途: 検出
難易度: Hard
コスト: Medium

Compositional Generalization via Structural Identification in a Category-Theoretic Framework

Compositional generalization is usually evaluated through model accuracy. We instead ask which structural or l

強化学習方策勾配 (PPO / A3C)テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Muon with Finite Newton-Schulz: The Smoothing Benefit in Nonsmooth Nonconvex Optimization

Muon has emerged as a strong optimizer for the matrix-valued parameters in large language model pretraining, a

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Fine-Tuning Whisper for Automatic Speech Recognition in Baniwa: A Preliminary Study

Automatic Speech Recognition (ASR) technologies have achieved remarkable performance in recent years through t

自然言語処理ファインチューニング分類テキスト音声

用途: 分類
難易度: Hard
コスト: Low

説明可能コンピュータビジョンセグメンテーションテキスト

arxivGitHubあり2026-08-26

Quantum-Inspired Modeling of Driving Behavior

Driver behavior is heterogeneous, context-dependent, and changes over time, and these properties shape the tra

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

説明可能深層学習軽量化・量子化検出異常検知テキスト

Efficient Estimation of High Information Projections using Nearest Neighbours

An intuitive method for dimensionality reduction is proposed, which is highly effective for finding interestin

用途: 検出
難易度: Hard
コスト: Medium

品質予測/異常検知自然言語処理大規模言語モデル生成テキスト

Homo-RAG: Homology-Guided Retrieval-Augmented Generation for Cross-Species Gene Function Prediction

The functional annotation of genes in non-model organisms remains a significant challenge in computational bio

用途: 生成
難易度: Hard
コスト: High

arxivGitHubあり2026-08-26

ES-AHD: An Evolution Strategy Framework for Automatic Heuristic Design

In this paper, we introduce ES-AHD, a novel framework that fundamentally integrates Evolution Strategy (ES) in

用途: 生成
難易度: Hard
コスト: High

Skill Issue: Are Skills Language-Invariant in LLMs?

Large language models access knowledge inconsistently across languages, but to what extent do they differ in t

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-08-25

Trust, but Verify: Rigorously Profiling Best-Effort High-Performance Computing for Digital Evolution

Developments in high-performance computing (HPC) technology continue to drastically increase quantities of ava

CPUで試しやすい品質予測/異常検知自然言語処理RAG生成テキスト

用途: 生成
難易度: Hard
コスト: Low

arxivPaper only2026-08-24

Integer Natural Evolution Strategies

While contemporary Evolution Strategies handle integer optimization problems effectively, their adaptation mec

強化学習マルチエージェントテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

arxivPaper only2026-08-24

FormuEvo: LLM-Guided Evolution for Discovering Solver-Efficient Mixed-Integer Programming Formulations

Mixed-integer programming (MIP) lies at the core of operations research and industrial optimization. While lar

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

条件最適化自然言語処理ファインチューニング生成回帰テキスト

arxivGitHubあり2026-08-24

JANUS: Online Jacobian-Aligned Infill for Black-Box Optimization

Population optimizers such as CMA-ES, DE, and multi-objective evolutionary algorithms drive search mainly thro

用途: 生成
難易度: Hard
コスト: High

arxivPaper only2026-08-22

Metric Distortion of Social Welfare Functions

Metric distortion has primarily been studied for social choice functions, which select a single winner from or

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

arxivPaper only2026-08-21

Event-triggered Implicit Perturbation for Zeroth-Order Fine-Tuning of Spiking Transformers

Zeroth-order (ZO) optimization estimates gradients using only forward-pass evaluations, making it suitable for

用途: 生成
難易度: Hard
コスト: Medium

arxivPaper only2026-08-20

The Forward-Backward Disconnect: State Dynamics, Credit Assignment, and Biological Grounding in Neural Computation

A recurring pattern in neural computation is the reintroduction of dynamical and biological structure into mod

説明可能深層学習RNN / LSTMテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-08-19

Graphical Design of Interpretable Architectures

Designing, implementing, and comparing interpretable architectures requires a formal language to represent the

説明可能数学・理論確率・統計テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

arxivPaper only2026-08-19

Beyond receptive fields: sequence-pooled normalization can supply most of a sequence labeler's context

A convolutional sequence labeler's receptive field is routinely treated as the extent of the model's usable co

深層学習CNNテキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-08-19

Preference Reasoning under Indeterminacy in Large Language Models

As large language models evolve into decision-making agents, the ability to reason over preferences becomes fu

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-08-18

Procedural Content Metageneration via Program Search and Continual Abstraction Discovery

Large language models can generate executable programs, which makes it possible to search directly over proced

用途: 生成
難易度: Hard
コスト: High

arxivPaper only2026-08-18

Computational Prosopography across a Millennium: Mathematically Oriented Lineages Traced from the Fields Medalists

We reconstruct the mentor--student network through which documented scholarly training passed across roughly n

用途: 生成
難易度: Hard
コスト: High

arxivPaper only2026-08-18

Planning Against Learning in Rank-1 Games

Learning algorithms are often used to make decisions in repeated multi-agent environments. When another player

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

arxivPaper only2026-08-17

Automating Parent Selection Configuration in Genetic Programming with Agentic AI

We investigate whether agentic artificial intelligence can automate parts of the process of designing genetic

自然言語処理大規模言語モデル生成回帰テキスト

用途: 生成
難易度: Hard
コスト: High

arxivPaper only2026-08-17

Evolving Executable Pipeline Programs for AutoML with Language Models

Automated machine learning (AutoML) systems search for pipelines within a space of preprocessing operators, le

表形式向きCPUで試しやすい自然言語処理大規模言語モデル分類テキスト表形式

用途: 分類
難易度: Hard
コスト: High

arxivPaper only2026-08-17

$\texttt{Flip-Team}$: Cooperative Takeover Games with Stochastic Human Override

Shared autonomy requires principled mechanisms for allocating and transferring control between a human and an

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

arxivPaper only2026-08-16

Mental Model Management: An Operator-Based Framework for LLM Memory

Large language models process large amounts of information but usually lack an explicit mechanism for maintain

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-08-16

Solvable Sokoban Without a Solver via Diffusion

Deciding whether a Sokoban puzzle is solvable is PSPACE-complete (Culberson, 1997): solutions can be exponenti

用途: 生成
難易度: Hard
コスト: High

arxivPaper only2026-08-16

When Search Eats the Web: A Model of Corpus Erosion under Generative Extraction

Generative search engines (GSEs) answer user queries directly from crawled web content. The capture of value f

品質予測/異常検知深層学習Transformer生成テキスト

用途: 生成
難易度: Hard
コスト: Low

arxivPaper only2026-08-15

Language models suffer from a curse of ambiguity

Large language models increasingly rely on sampling as a driver of their own improvement, making the fidelity

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-08-15

Chameleon: An Adaptive AI-Driven Honeypot Architecture Using Threat-Calibrated Particle Swarm Optimization and Semantic Deception Rapidly-Exploring Random Trees

An invariant behavioral profile is the defining vulnerability of traditional honeypot installations: a skilled

CPUで試しやすい深層学習RNN / LSTM分類生成テキスト

用途: 分類
難易度: Hard
コスト: Low

arxivPaper only2026-08-15

Watermarked Game Solving via Perturbed Regret Minimization

Many real-world interactions among self-interested parties can be modeled by game theory, and the rapid advanc

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-08-14

The Open-Strategy Dictator Game: Cooperation Under Mutual Transparency

We introduce the Open-Strategy Dictator Game (OSDG), a variant of the classic dictator game in which each play

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Beyond the Best Guess: Improving LLM Solution Coverage with Evolution Strategies

解決策候補の多様化を向上させる進化戦略を開発するために、LLMを用いて解決策候補を生成するアプローチを提案している。

自然言語処理大規模言語モデルテキスト強化学習

用途: 解決策候補の多様化を向上させる進化戦略
難易度: Hard
コスト: High

Keep, Customize, or Exit: Default Design and Token Pricing in LLM Reasoning Services

この作品は、Large Language Model (LLM) サービスにおけるデフォルト設計とトークンの価格設定を行う為の機械学習手法を提案しています。この手法は、LLM サービスにおけるデフォルト設計とトークンの価

用途: LLMサービスにおけるデフォルト設計とトークンの価格設定
難易度: Hard
コスト: High

Power in Liquid Democracy: A Network Centrality Approach

Liquid民主主義の意思決定において、意思決定ネットワークの中断度を考慮した力関係を計算方法を提案。意思決定の力関係を計測し、意思決定の透明性と責任性を高める。

強化学習方策勾配 (PPO / A3C)テキスト

用途: Liquid民主主義の意思決定の力関係
難易度: Hard
コスト: Medium

Error-Aware Reverse Auction Mechanism for Large Language Model Routing

大規模言語モデルによる質問のルーティングをオープン化するために、リバーサルオークションを用いて質問のルーティングを実現する。

用途: 大規模言語モデルによる質問のルーティングをオープン化
難易度: Hard
コスト: High

arxivPaper only2026-08-12

Large Language Models as Strategic Bidding Agents in P2P Energy Trading Markets

Peer-to-peer (P2P) energy trading markets rely on double auction mechanisms to match prosumers and consumers i

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

品質予測/異常検知強化学習方策勾配 (PPO / A3C)テキスト

Contextual Quality-Diversity Evolutionary Reinforcement Learning for HVAC Control in Tropical Commercial Buildings

この研究では、熱帯気候の商業ビルでの冷房システムの制御に適した、コンテキストベースの品質多様性進化的強化学習を提案します。制御システムは、データドライブの稼働状況、日々の気象と負荷のシナリオ、コンテキスト無関係の行動記述

用途: HVACシステムの制御
難易度: Hard
コスト: Medium

深層学習Transformer分類セグメンテーション画像

Spiking Local Interaction and Adaptive Complementary Fusion for Spiking Transformer

Spiking Transformers model token interactions primarily through spiking self-attention (SSA). However, binary

用途: 分類
難易度: Hard
コスト: High

EvoMem: Memory-Augmented Evolution for Code Optimization

成功した突然変異戦略の中には、単一の実行で利用可能な知識が存在し、その知識は複数のタスク間で移行することができる。しかし、既存のLLM-ベースの進化的フレームワークでは、再利用可能な知識は捨てられ、同じアイディアの再発見

自然言語処理大規模言語モデルQAテキスト

用途: コード最適化問題解決
難易度: Hard
コスト: High

Persistent Recursive Worlds Enable Autonomous Software Evolution

この研究では、ソフトウェアの開発が複数のエージェントによって長期間にわたって進行する場合の持続可能性を考慮した新しいアプローチであるEvoX Genesisを提案します。

生成AIGANテキスト

用途: ソフトウェア開発の持続可能性の向上
難易度: Hard
コスト: Medium

説明可能自然言語処理大規模言語モデルテキストマルチモーダル

VERDICT: Training-Free Step-Wise Verification of Multimodal Reasoning via Disagreement-Aware Consensus

VERDICT は、マルチモーダル論理の確認と検証をサポートするフレームワークである。このフレームワークでは、多くの場合、確認と検証は、エージェントが提供するさまざまなスコアを合計すると考えられてきたが、このフレームワー

用途: マルチモーダル論理の確認と検証
難易度: Hard
コスト: High

Evaluating Rational Contracting in Natural Language

The emergence of language-based AI agents promises to transform the scope of machine economic activity. Instea

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

The Game of Marginal Utilities

We study a noncooperative resource-allocation game in which $m$ players distribute fixed resources among $n$ p

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Neuroevolution Arena: Nested Ecological Evaluation of Update-and-Inheritance Regimes across Neural Architectures

Competitive artificial-life systems can rank trained controllers differently under training and ecological eva

用途: 生成
難易度: Hard
コスト: High

BDH-CQ: In-Context Learning with Recurrent Latent Reasoning

この研究では、BDH-CQ（Bert Discriminator-Hidden Chain Question）と呼ばれる新しい認識モデルを提案した。これは、従来の言語モデルに加えて、リーソン（reasoning）機能を持

深層学習RNN / LSTMテキスト

用途: リーソンを実現するための新しい認識モデル
難易度: Hard
コスト: Medium

Identifying potentiating events in evolutionary search using replay experiments

In this work, we introduce analytical replay experiments to the evolutionary computing community. Replay exper

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Resolving Envy by Adding Goods with Bounded Supply: A Type-Count Dichotomy and Two-Agent Hardness

We study envy elimination by adding goods (EEAG) when the additional pool has bounded supply and no separate b

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

Competitive mediator games and urban CAV routing markets

Inspired by possible future markets of autonomous routing and driving (ARAD), we introduce competitive mediato

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

Regret, equilibrium, and learning in games: A guided tour

This note aims to serve as an entry point to the literature on learning in games, a topic with significant the

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

arxivPaper only2026-08-09

Complete EFX Allocations Exist for Four Additive Agents and Up to Nine Goods

We prove that every fair-division instance with four agents, additive valuations over the non-negative reals,

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

arxivGitHubあり2026-08-08

A Hybrid Nested Harness for Decoupling Structure and Parameters in LLM-Driven Optimization

In evolutionary algorithms powered by language models, the LLM acts as a single operator that simultaneously u

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-08-08

RotaryQuant: Fitting 120B MoE Models on Consumer Hardware via Fused Compressed-Space Attention

Large mixture-of-experts (MoE) language models with 26--120 billion parameters exceed the memory capacity of c

深層学習正規化・最適化手法テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Low

arxivPaper only2026-08-07

Recipes for Creativity: Iterative Generation and Evaluation in Large Language Models

Generative models are often evaluated through singular artifacts, whereas human creativity typically emerges t

用途: 生成
難易度: Hard
コスト: High

arxivPaper only2026-08-06

Relay, Don't Route: Adaptive Population Handoff for Cost-Efficient LLM-Driven Evolution

Relayは、計算コストを削減しながらLLMエボリューションのための並列プロセスの一時的なハンドオフを実現します。

用途: LLMドライブエボリューション
難易度: Hard
コスト: High

arxivPaper only2026-08-05

Emergence of Reputation-Based Cooperation in LLM Agents

Can cooperation among large language model (LLM) agents be evolutionarily stable against free-rider invasion?

深層学習Transformer生成画像テキスト

用途: 生成
難易度: Hard
コスト: High

The Transformer Revolution, Part 1: Dynamic Processing through Output-Weight Interconnections

作品たちは、大規模言語モデルによって生成された結果の理解を進めます。大規模言語モデルは、概念を生成するための新しいメカニズムを提供します。

説明可能深層学習Transformerテキスト

用途: 生成モデルによる理解
難易度: Hard
コスト: High

Omega-S: A Functional Resilience Index for LLM Fine-Tuning

LLMが外部アクションを取り続けている場合、エージェントのメモリーが古くなったり誤った情報を持ったりする可能性があります。この問題を解決するために、この研究ではSafeCommitという技術を提案しています。SafeCo

深層学習正規化・最適化手法テキスト

用途: LLMによるメモリー・グランド・エージェントの安全なアクション
難易度: Hard
コスト: High

MuEvo: LLM-Driven Evolution of Multi-Heuristic Ensemble

この研究では、LLMを用いてマルチハイスティックエンサンブルを進化させる技術を提案するものです。エンサンブルの各ハイスティックは単独で効果的なものではなく、他のハイスティックと協力して最高の結果を達成するものと想定され

用途: マルチハイスティックエンサンブルの進化
難易度: Hard
コスト: High

NeuroMosaic: Anatomically Grounded Multimodal Large Language Modeling for Molecularly Aware Glioma Reasoning from 3D MRI and Clinical Narratives

この研究では、3D MRIと臨床記録を活用した大規模言語モデルの開発を提唱。提案されたNeuroMosaicは、医療画像を解剖学的情報に基づく地域情報に変換し、臨床記録と分子情報に調整を行い、MRI領域との接続を確実にし

自然言語処理大規模言語モデル画像テキスト3D

用途: 多様な医療画像と臨床記録からがんの推定を実現
難易度: Hard
コスト: High

arxivPaper only2026-08-03

ChaosProbe: A Neurochaotic Lens on Frozen Transformer Input-Embedding Spaces

フリーザントランソーメールの特性化を改善し、非線形のプローブを使用した。

品質予測/異常検知深層学習Transformer生成テキスト

用途: フリーザントランソーメールの特性化
難易度: Hard
コスト: High

arxivPaper only2026-08-03

Towards Autonomous Formulaic Alpha Discovery: An Evolutionary Computation Perspective

Automated formulaic alpha discovery aims to generate predictive and interpretable trading signals from large s

説明可能品質予測/異常検知自然言語処理大規模言語モデル生成テキスト強化学習

用途: 生成
難易度: Hard
コスト: High

arxivGitHubあり2026-08-03

Spike-HTR: Spiking Neural Transformer for Handwritten Text Recognition

Handwritten Text Recognition (HTR) is computationally imbalanced in two ways: most image pixels are background

深層学習Transformer分類画像テキスト

用途: 分類
難易度: Easy
コスト: Low

arxivPaper only2026-08-02

Unsupervised Multidomain Approaches to Named Entity Recognition with Small Datasets

This paper explores the challenges and the methodologies associated with learning quality representations in s

少数データ向き品質予測/異常検知深層学習軽量化・量子化分類テキスト教師なし

用途: 分類
難易度: Hard
コスト: High

arxivPaper only2026-08-01

DGA$_2$D: Directed Graph-Guided Automated Algorithm Design with Large Language Models

The rapid development of Large Language Models (LLMs) has opened new avenues for Automated Heuristic Design (A

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-07-31

MOSAIC: Adversarial Co-evolution of Specialist Heuristics and Problem Instances for LLM-based Automated Heuristic Design

Automated heuristic design (AHD) with large language models (LLMs) has produced strong heuristics for combinat

品質予測/異常検知自然言語処理大規模言語モデル生成テキスト

用途: 生成
難易度: Hard
コスト: High

arxivPaper only2026-07-31

SILVA Networks as Structured Implicit Layers and Vector Attractors via Dynamic Interaction Fields

Many learning problems require representations that reconcile direct input, nearby structure, and broader cont

コンピュータビジョンセグメンテーション分類画像テキスト

用途: 分類
難易度: Hard
コスト: Low

arxivPaper only2026-07-30

Nanoparticle Networks for Neuromorphic Computing

ニュロモーティックコンピューティングの研究を目的としたフレームワーク

用途: ニュロモーティックコンピューティング
難易度: Hard
コスト: Low

arxivPaper only2026-07-30

Guiding Large Language Models with Genetic Programming-Evolved Heuristic Knowledge for Dynamic Multi-Mode Project Scheduling

In dynamic multi-mode project scheduling, activities have alternative execution modes and uncertain durations,

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-07-29

The Sparsity Ceiling: Where Spiking Networks Can and Cannot Trade Activity for Energy

Spiking neural networks (SNNs) are promoted as an energy-efficient substrate because sparse, event-driven acti

用途: 技術検証・論文読解補助
難易度: Hard
コスト: Medium

arxivPaper only2026-07-29

EvoPINN: Agentic Discovery of Executable Algorithms for Physics-Informed Neural Networks

物理的システムの分離方程式を解くためには、ニュラルネットワークの設計、損失関数の定義、および最適化ダイナミクスの手動調整が必要である。研究者は、自動設計のためにLarge Language Models (LLMs)を利

用途: 物理的システムの分離方程式を解く
難易度: Hard
コスト: High

arxivPaper only2026-07-29

An evolutionary model of animats with VLM-based subjective evaluation

In this study, we propose a framework that incorporates subjective evaluations provided by a Vision-Language M

コンピュータビジョンマルチモーダル画像テキスト

用途: 技術検証・論文読解補助
難易度: Hard
コスト: High

arxivPaper only2026-07-26

Benchmarking Zero-Shot LLM-Generated Parent Selection in Genetic Programming for Symbolic Regression

Parent selection significantly affects exploration, exploitation, and complexity control in genetic programmin

自然言語処理大規模言語モデル生成回帰テキスト

用途: 生成
難易度: Hard
コスト: High

arxivPaper only2026-07-25

Continuous surrogates versus threshold Boolean networks for modeling Arabidopsis ISR gene regulation

Gene regulatory network modeling often requires balancing predictive accuracy and mechanistic interpretability

表形式向き説明可能CPUで試しやすい自然言語処理RAG回帰テキスト

用途: 回帰
難易度: Hard
コスト: Low

arxivPaper only2026-07-22

The Giant Hippocampus: From Structural Monoculture to a System of Systems

この研究では、人工知能の研究者と神経科学者の間の分野を結びつけるために、脳のシステム構造を研究し、その研究から導かれた新しいアプローチを提案しました。

深層学習Transformer分類画像テキスト

用途: 脳のシステム構造とその応用
難易度: Hard
コスト: High