「Calibration」の検索結果
77 件Standard spatiotemporal graph convolutional networks need a fixed-size adjacency matrix, coupling model weight
Vision-language models (VLMs) increasingly support construction hazard monitoring, yet their frequent false al
Power-grid customer service increasingly relies on community social media, where customers describe outages, b
Semi-supervised medical image segmentation aims to train reliable segmentation models using limited labeled da
Conformal prediction (CP) provides marginal coverage guarantees, but marginal coverage does not determine wher
Learn-then-Test (LTT) provides a general framework for certifying deployment policies against a prescribed ris
We study score learning from denoising pairs when the data contain a hidden non-Gaussian component. If covaria
Group-based reinforcement learning with verifiable rewards (RLVR) learns only from prompts whose sampled compl
Time series foundation models (TSFMs) are increasingly used for probabilistic forecasting across diverse domai
As control engineering is applied to increasingly complex systems, learning-based control, which uses empirica
Suicide prevention remains a critical public health challenge, and electronic health records (EHRs) provide an
Generative AI is billed per token, yet tokens are not equally expensive. Each token enters the KV-cache and ra
Calibrated multiaccuracy asks that a predictor be globally calibrated and multiaccurate with respect to a boun
As general-purpose Large Language Models (LLMs), such as Gemini and DeepSeek, exhibit increasingly advanced re
An aggregate task score does not distinguish failures of scheduling, state retention, and candidate selection.
Recourse-aware learning trains models so that people who receive an adverse decision can reverse it with an af
Laminar-turbulent transition governs skin friction, heat transfer and losses in turbomachinery, yet Reynolds-a
LLM post-training via reinforcement learning has shown consistent improvements in accuracy at the cost of redu
Financial workflows need fast, calibrated, auditable decisions, not just generated text. We train System-1 dec
Fusing independently normalized attention heads can increase the number of explicit key-value entries, or atom
We compare three zero-shot time-series foundation models (Timestral-5M, Timestral-209M and Chronos-2, 8192-hou
Best-of-N decoding improves generation quality by sampling multiple candidates and selecting the best, but it
Large language model (LLM) embeddings have been increasingly adopted in sequential recommendation to encode st
Kolmogorov-Arnold Networks (KANs) have recently been proposed as a promising alternative to Multi-Layer Percep
Sampling and selection spend test-time compute inside a candidate pool, but they cannot recover a correct answ
Programs that use a language model as a classifier act on its most probable label. Text injected into the inpu
Retrieval-augmented generation (RAG) extends large language models with external evidence, yet relevant eviden
Earth observation (EO) workflows require selecting among known outcomes, estimating confidence, and processing
Multimodal test-time adaptation (TTA) enables pretrained multimodal models to accommodate distribution shifts
Accurate full-body tracking in Virtual Reality (VR) is heavily underconstrained when relying solely on consume
Financial document recognition is commonly evaluated using character and word error rates, which do not adequa
3-D road anomaly segmentation is essential for risk-aware perception in autonomous driving. Its central challe
Myoelectric Control Systems (MCSs) enable voluntary control of wearable robots through muscle activity. Howeve
Vision-language-action (VLA) models offer a promising path toward general-purpose robot manipulation. While in
Markerless hand-eye calibration under continuous motion requires jointly resolving camera geometry and the tim
Recent robot systems combine wrist and fixed cameras, requiring camera-to-robot transforms for both. Most hand
Online large language model (LLM) services face a pricing cold start: providers initially lack the historical
Training a classifier means letting a validation metric decide which model is kept and when training stops. A
Tensor factor models provide low-dimensional representations of tensor-valued observations while preserving th
Fairness-aware classification seeks to balance predictive accuracy and fairness across demographic groups. Alt
Selective prediction with split conformal calibration promises a post-hoc fix for molecular property predictio
Uncertainty estimates from fuzzy regression are informative only when the implementation is correct and the co
Key-value (KV) cache eviction methods rank tokens by attention scores, and each layer scores its own cache. To
Bursty serving changes both queue states and the population selected for each model. We study conditional cali
On-policy distillation (OPD) trains a student on its own generations by matching the teacher's token probabili
Multiple studies have shown that neural networks can be behaviorally systematic, producing correct outputs for
Automated classification of retinal pathologies from Optical Coherence Tomography (OCT) images remains challen
We study simultaneous quantile inference under local differential privacy for a number of target probabilities
Interaction trees reuse outcomes to discover treatment-effect heterogeneity, complicating inference after grow
Machine learning can screen far more candidates than can be experimentally validated, creating a need for sele
Accurate posture feedback is essential for precise control of continuum robots. Off-the-shelf bend sensors mea
Voluntary platoon formation requires vehicles to coordinate despite uncertain downstream fuel savings and priv
Data-driven learning for the solution operator of a differential equation has broad scientific importance, but
Modern data analysis pipelines increasingly rely on large datasets assembled through retrieval, scraping, auto
Artificial intelligence (AI) models can accelerate cell-type annotation in digital pathology, but safely scali
Adaptive inference systems balance accuracy and inference cost by routing inputs to models with varying capabi
We study adaptive UAV sampling for radio environment map (REM) construction from non-Gaussian IoT observations
Rotary Position Embedding (RoPE) typically uses uniform coverage across Transformer layers despite heterogeneo
Parameter estimation and classification in binary logistic regression deteriorate severely when the number of
Proactive agents must decide whether, when, and how far to assist without an explicit request. Existing evalua
Learning independence is fundamental to statistical inference and machine learning, yet learning independence
Ship-to-home fulfillment requires a retailer to select a fulfillment location and carrier before the delivery
We study conformal prediction under covariate shift with a known covariate density ratio, seeking prediction s
Objective: Variable selection for clinical prediction models is commonly done on statistical grounds related t
Variational inference with normalizing flows provides a flexible posterior approximation for Bayesian model ca
Model discrepancy may prevent a single calibration parameter from reproducing a physical response across the c
Nonlinear structural system identification remains challenging when governing physics is only partially known
The quality with which language models follow instructions falls as the load grows, yet there is still no quan
In large language model (LLM)-based multi-agent systems, reported confidence increasingly determines the alloc
Event-based spiking neural networks (SNNs) encode information through temporal firing dynamics, but the same s
Command-line interface (CLI) agents increasingly execute longer-horizon tasks, yet existing benchmarks either
Closed-loop neuromodulation requires sleep staging that respects stage-specific minimum dwells: stimulation po
Opponent histories can mislead after behavioral shifts. SafeAdapt-Poker formulates budgeted adaptation under o
Fast and accurate qubit-state assignment is essential for feedback, calibration, and error correction in quant
Low-cost NAS must both explore high-performing architectures and identify them reliably, yet reducing evaluati
Randomised copies of a connectome are the usual baseline for asking whether measured wiring matters, and the a
This work evaluates the direct transfer of a co-evolved communication protocol from a 2D simulation to a 3D ph