MLinfo | 機械学習・AI論文まとめ

On the Geometry of On-Policy Distillation

On-policy distillation (OPD) is increasingly used to improve large language model reasoning, but its training

深層学習軽量化・量子化検出生成テキスト

用途: 検出
難易度: Easy
コスト: High

DuMate-DeepResearch: An Auditable Multi-Agent System with Recursive Search and Rubric-Grounded Reasoning

Deep Research (DR) has emerged as a new agentic paradigm to tackle complex, open-ended research tasks, demandi

品質予測/異常検知強化学習マルチエージェント生成

用途: 生成
難易度: Easy
コスト: Low

Where Rectified Flows Leak: Characterising Membership Signals Along the Interpolation Path

Understanding what generative models retain from training data remains challenging, with implications for copy

機械学習特徴量エンジニアリング生成画像音声

用途: 生成
難易度: Easy
コスト: High

MI向き自然言語処理大規模言語モデル生成画像テキスト

MMAE: A Massive Multitask Audio Editing Benchmark

We introduce MMAE, a Massive Multitask Audio Editing benchmark, serving as the first comprehensive evaluation

用途: 生成
難易度: Easy
コスト: High

UniSHARP: Universal Sharp Monocular View Synthesis

In this work, we focus on extending SHARP, the popular photorealistic view synthesis method, for universal mon

コンピュータビジョン3D・点群生成画像3D

用途: 生成
難易度: Easy
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・量子化生成テキスト音声

dots.tts Technical Report

We present dots.tts, a 2B-parameter continuous autoregressive text-to-speech (TTS) foundation model that model

用途: 生成
難易度: Easy
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

Socratic-SWE: Self-Evolving Coding Agents via Trace-Derived Agent Skills

LLM-driven software engineering agents have become a central testbed for real-world language-model capability,

用途: 生成
難易度: Easy
コスト: High

Stream3D-VLM: Online 3D Spatial Understanding with Incremental Geometry Priors

Despite advances in 3D scene understanding, existing 3D Large Multimodal Models operate in offline settings, r

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: Easy
コスト: High

センサ/時系列自然言語処理ファインチューニング生成テキスト音声

Entropy as a Structural Prior: How a Log-Barrier on DiT Belief Space Drives Musical Diversity and Development

Confidence-based loss weighting is usually avoided in generative models because it accelerates errors when the

用途: 生成
難易度: Easy
コスト: High

Empirical Study on the Characteristics and Evolution of AI-usage in GitHub Repositories: Evidence from Code Comments

Developers increasingly use AI tools such as ChatGPT, Copilot, and Claude in everyday software workflows, but

深層学習Transformer分類生成テキスト

用途: 分類
難易度: Easy
コスト: High

Streaming Video Generation with Streaming Force Control

We introduce StreamForce, a streaming video generation framework that enables physically grounded control thro

深層学習軽量化・量子化生成動画

用途: 生成
難易度: Easy
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

Cosine Misleads: Auxiliary Losses Reshape Vision Language Models, Not Their Latents

Latent visual reasoning (LVR) inserts supervised latent tokens between perception and answer generation in vis

用途: 生成
難易度: Easy
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3D・点群生成画像テキスト

Direct 3D-Aware Object Insertion via Decomposed Visual Proxies

Object insertion aims to seamlessly composite a reference object into a specified region of a background image

用途: 生成
難易度: Easy
コスト: High

Physics in 2-Steps: Locking Motion Priors Before Visual Refinement Erases Them

Image-to-Video diffusion models leverage input images to generate visually stunning content, yet frequently pr

自然言語処理RAG生成画像動画

用途: 生成
難易度: Easy
コスト: High

huggingfaceGitHubありHugging Faceあり2026-06-04

Almieyar-Oryx-BloomBench: A Bilingual Multimodal Benchmark for Cognitively Informed Evaluation of Vision-Language Models

Despite the rapid progress of Vision-Language Models (VLMs), the field lacks benchmarks that rigorously diagno

品質予測/異常検知深層学習Transformer生成画像テキスト

用途: 生成
難易度: Easy
コスト: High

LoomVideo: Unifying Multimodal Inputs into Video Generation and Editing

Developing unified video generation and editing models capable of interpreting interleaved multimodal inputs i

深層学習Transformer生成画像テキスト

用途: 生成
難易度: Easy
コスト: High

huggingfaceGitHubありHugging Faceあり2026-06-04

MLEvolve: A Self-Evolving Framework for Automated Machine Learning Algorithm Discovery

Large language model (LLM) agents are increasingly applied to long-horizon tasks such as scientific discovery

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: Easy
コスト: High

huggingfaceGitHubありHugging Faceあり2026-06-04

Dream.exe: Can Video Generation Models Dream Executable Robot Manipulation?

Video generation models have made impressive strides in synthesizing visually compelling content, yet their ou

品質予測/異常検知自然言語処理RAG生成画像動画

用途: 生成
難易度: Easy
コスト: High

Complexity-Balanced Diffusion Splitting

Standard continuous-time generative models rely on monolithic architectures that must navigate vastly differen

品質予測/異常検知深層学習軽量化・量子化生成

用途: 生成
難易度: Easy
コスト: High

品質予測/異常検知深層学習軽量化・量子化生成教師なし

Unsupervised Skill Discovery for Agentic Data Analysis

Inference-time skill augmentation provides a lightweight way to improve data-analytic agents by injecting reus

用途: 生成
難易度: Easy
コスト: Low

LLMs Can Leak Training Data But Do They Want To? A Propensity-Aware Evaluation of Memorization in LLMs

Large language models can reproduce training data, but existing memorization evaluations mostly measure whethe

深層学習軽量化・量子化生成テキスト

用途: 生成
難易度: Easy
コスト: High

Latent Reasoning with Normalizing Flows

Large language models often improve reasoning by generating explicit chain-of-thought (CoT), demonstrating the

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: Easy
コスト: High

Towards Truly Multilingual ASR: Generalizing Code-Switching ASR to Unseen Language Pairs

Automatic Speech Recognition (ASR) has become a key technology for human--AI interaction. However, code-switch

自然言語処理ファインチューニング分類生成音声

用途: 分類
難易度: Easy
コスト: Low

Discrete-WAM: Unified Discrete Vision-Action Token Editing for World-Policy Learning

Autonomous driving requires reasoning about how ego actions shape the evolution of the surrounding world. Howe

生成AI拡散モデル生成画像

用途: 生成
難易度: Easy
コスト: High

ForeSci: Evaluating LLM Agents for Forward-Looking AI Research Judgment

AI research often requires decisions before future evidence exists: which bottleneck to attack, which directio

自然言語処理大規模言語モデル生成予測

用途: 生成
難易度: Easy
コスト: High

VideoKR: Towards Knowledge- and Reasoning-Intensive Video Understanding

We introduce VideoKR, the first large-scale training corpus specifically designed to strengthen knowledge- and

自然言語処理ファインチューニング生成テキスト動画

用途: 生成
難易度: Easy
コスト: High

SePO: Self-Evolving Prompt Agent for System Prompt Optimization

System prompt optimization improves agent behavior without modifying the underlying model, yielding human-read

自然言語処理RAG生成テキスト

用途: 生成
難易度: Easy
コスト: High

BRepCLIP: Contrastive Multimodal Pretraining on BRep Primitives for CAD Understanding

Learning representations of CAD models is a largely open problem. While 3D representation learning has flouris

深層学習Transformer分類生成埋め込み

用途: 分類
難易度: Easy
コスト: High

品質予測/異常検知深層学習Transformer生成動画

Echo-Infinity: Learning Evolving Memory for Real-Time Infinite Video Generation

We present Echo Infinity, an autoregressive (AR) framework towards real-time infinite video generation that em

用途: 生成
難易度: Easy
コスト: High

センサ/時系列コンピュータビジョンマルチモーダル生成画像

MapAgent: An Industrial-Grade Agentic Framework for City-scale Lane-level Map Generation

Lane-level maps are critical infrastructure for autonomous driving and lane-level navigation, yet constructing

用途: 生成
難易度: Easy
コスト: High

GRAIL: Generating Humanoid Loco-Manipulation from 3D Assets and Video Priors

Scaling humanoid loco-manipulation requires robot-compatible demonstrations across diverse objects, whole-body

コンピュータビジョン3D・点群生成画像動画

用途: 生成
難易度: Easy
コスト: High

MeshWeaver: Sparse-Voxel-Guided Surface Weaving for Autoregressive Mesh Generation

Autoregressive mesh generation has gained attention by tokenizing meshes into sequences and training models in

深層学習Attention機構生成テキスト3D

用途: 生成
難易度: Easy
コスト: High

huggingfaceGitHubありHugging Faceあり2026-06-03

SpeechEditBench: A Bilingual Multi-Attribute Benchmark for Instruction-Guided Speech Editing

Instruction-guided speech editing requires a model to modify specified speech attributes while preserving unre

自然言語処理大規模言語モデル生成テキスト音声

用途: 生成
難易度: Easy
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

Text-to-Image Models Need Less from Text Encoders Than You Think

Text-to-image models rely on text prompts as their primary interface to human intent. Prompts are encoded by a

用途: 生成
難易度: Easy
コスト: High

A Cookbook of 3D Vision: Data, Learning Paradigms, and Application

3D vision has rapidly evolved, driven by increasingly diverse data representations, learning paradigms, and mo

自然言語処理RAG生成動画3D

用途: 生成
難易度: Easy
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

RobotValues: Evaluating Household Robots When Human Values Conflict

While household robots are often evaluated based on task completion, everyday domestic environments involve va

用途: 生成
難易度: Easy
コスト: High

MI向き自然言語処理RAG生成セグメンテーション画像

MAOAM: Unified Object and Material Selection with Vision-Language Models

Selection is a core operation in interactive image editing. To be practical, a user should be able to specify

用途: 生成
難易度: Easy
コスト: High

Qwen-Image-Flash: Beyond Objective Design

Few-step distillation has become an effective strategy for accelerating advanced visual generative models, yet

MI向き深層学習軽量化・量子化生成画像テキスト

用途: 生成
難易度: Easy
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動画

OVO-S-Bench: A Hierarchical Benchmark for Streaming Spatial Intelligence in Multimodal LLMs

Multimodal agents in robotics, AR, and autonomous driving must reason about places and layouts from continuous

用途: 生成
難易度: Easy
コスト: High

huggingfaceGitHubありHugging Faceあり2026-06-02

Self-Distilled Policy Gradient

On-policy self-distillation, where a language model conditions on privileged context to supervise its own gene

深層学習軽量化・量子化生成テキスト強化学習

用途: 生成
難易度: Easy
コスト: Medium

Eliciting Complex Spatial Reasoning in MLLMs through Wide-Baseline Matching

Wide-baseline matching (WBM) requires integrating geometric understanding, viewpoint changes, fine-grained per

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: Easy
コスト: High

AAD-1: Asymmetric Adversarial Distillation for One-Step Autoregressive Video Generation

We present AAD-1, an Asymmetric Adversarial Distillation framework for One-step autoregressive image-to-video

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: Easy
コスト: High

Large Language Models Hack Rewards, and Society

Reinforcement learning (RL) has become a dominant post-training paradigm, enabling large language models (LLMs

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: Easy
コスト: High

BraveGuard: From Open-World Threats to Safer Computer-Use Agents

Computer-use agents extend language models from text generation to sustained interaction with files, terminals

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: Easy
コスト: High

huggingfaceGitHubありHugging Faceあり2026-06-02

Agentic Chain-of-Thought Steering for Efficient and Controllable LLM Reasoning

Large language models improve final-answer accuracy through extended chain-of-thought reasoning, but often spe

深層学習軽量化・量子化生成テキスト強化学習

用途: 生成
難易度: Easy
コスト: High

huggingfaceHugging Faceあり2026-06-01

The Road Ahead in Autonomous Driving: The KITScenes Multimodal Dataset

Existing autonomous driving datasets have enabled major progress, but fall short in sensor fidelity, map compl

センサ/時系列深層学習Transformer検出生成3D

用途: 検出
難易度: Easy
コスト: High

huggingfaceGitHubありHugging Faceあり2026-06-01

Cosmos 3: Omnimodal World Models for Physical AI

We introduce Cosmos 3, a family of omnimodal world models designed to jointly process and generate language, i

深層学習Transformer生成画像テキスト

用途: 生成
難易度: Easy
コスト: High

huggingfaceHugging Faceあり2026-06-01

Where Do Deep-Research Agents Go Wrong? Span-Level Error Localization in Agent Trajectories

Deep-research agents solve tasks through long trajectories of search, tool use, evidence inspection, and answe

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

用途: 検出
難易度: Easy
コスト: High

huggingfaceGitHubありHugging Faceあり2026-06-01

Scalable Inference-Time Annealing with Surrogate Likelihood Estimators

A long standing challenge in computational chemistry and biophysics is efficiently sampling the Boltzmann dist

説明可能深層学習軽量化・量子化生成

用途: 生成
難易度: Easy
コスト: High

huggingfaceHugging Faceあり2026-05-31

BenchEvolver: Frontier Task Synthesis via Solution-Centric Evolution

The rapid progress of frontier large language models has led to widespread benchmark saturation, limiting the

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: Easy
コスト: High

huggingfaceHugging Faceあり2026-05-30

SDR: Set-Distance Rewards for Radiology Report Generation

Reinforcement learning with verifiable rewards has rapidly advanced reasoning in vision--language models. Howe

品質予測/異常検知深層学習Transformer生成テキスト強化学習

用途: 生成
難易度: Easy
コスト: High

huggingfaceHugging Faceあり2026-05-29

Human Psychometric Questionnaires Mischaracterize LLM Behavior

We examine whether human psychometric questionnaires can serve as reliable tools for characterizing and predic

自然言語処理大規模言語モデル生成

用途: 生成
難易度: Easy
コスト: High

huggingfaceGitHubありHugging Faceあり2026-05-29

The Distillation Game: Adaptive Attacks & Efficient Defenses

Distillation attacks create a deployment trade-off for model providers: the same outputs that make a model mor

品質予測/異常検知深層学習軽量化・量子化生成

用途: 生成
難易度: Easy
コスト: Medium

huggingfaceHugging Faceあり2026-05-29

Combinatorial Synthesis: Scaling Code RLVR via Atomic Decomposition and Recombination

Reinforcement Learning with Verifiable Rewards (RLVR) has recently emerged as the cornerstone for shaping the

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: Easy
コスト: High

huggingfaceGitHubありHugging Faceあり2026-05-29

OpenSTBench: Beyond Semantic Evaluation for Speech Translation

Speech translation systems increasingly span speech-to-text translation (S2TT), speech-to-speech translation (

品質予測/異常検知コンピュータビジョン動画認識生成テキスト音声

用途: 生成
難易度: Easy
コスト: High

huggingfaceHugging Faceあり2026-05-29

PaintBench: Deterministic Evaluation of Precise Visual Editing

While current multimodal models are proficient at open-ended visual editing, executing precise single-answer e

コンピュータビジョンマルチモーダル生成画像

用途: 生成
難易度: Easy
コスト: High

huggingfaceGitHubありHugging Faceあり2026-05-29

Score-Control for Hallucination Reduction in Diffusion Models

Diffusion models have emerged as the backbone of modern generative AI, powering advances in vision, language,

コンピュータビジョンセグメンテーション生成画像音声

用途: 生成
難易度: Easy
コスト: High

huggingfaceHugging Faceあり2026-05-24

WorldCraft: From Camera Navigation to Object Manipulation in Interactive Video World Models

Recent video-based world models have made pixel-space environments interactive at the camera level: users can

自然言語処理ファインチューニング生成画像動画

用途: 生成
難易度: Easy
コスト: High

huggingfaceHugging Faceあり2026-05-04

Liberating LLM Capabilities in Full-Duplex Speech Models

Speech-based large language models are typically constrained to spoken replies, which limits their user-facing

自然言語処理大規模言語モデル生成テキスト音声

用途: 生成
難易度: Easy
コスト: High