CORTEXA
← Browse
arxivcs.LGcs.AI2026-06-25

PEBS: Per-rater Empirical-Bayes Shrinkage for RLHF Reward-Model Calibration

Arnav Raj

Reward models for Reinforcement Learning from Human Feedback (RLHF) pool preferences across thousands of annotators and fit one global affine calibrator, collapsing raters with systematically different rating-scale offsets and slopes into a single average-rater fit that does not match any individual annotator. PEBS is a per-rater empirical-Bayes shrinkage estimator: it fits per-rater affine calibrators on a held-out slice of each annotator's ratings and applies Morris-James-Stein empirical-Bayes shrinkage toward the population mean, in closed form and without retraining the reward model. On PRISM, PEBS reduces within-user held-out RMSE by 8.58% over the pooled population-slope baseline. The procedure replicates on PluriHarms harm ratings (Qwen-2.5 base, in-family) with a +9.66% RMSE reduction over the same population-slope baseline. PEBS is a closed-form post-hoc estimator for annotator-specific affine calibration in RLHF reward modeling; it leaves the reward base model unchanged and estimates only the rater-level map used at inference time for new ratings.

View free PDFSource page

Related papers

arxivstat.MLcs.AIcs.LGeess.SYstat.ME2026-06-25

XMSE-Aware Adaptive Empirical Bayes Estimation

Minghao Chen, Jiale Zheng

Empirical Bayes (EB) estimators can match the first-order asymptotic risk of maximum likelihood (ML) while behaving very differently at second order: recent excess mean squared error (XMSE) analysis shows that kernel-based EB estimation may be worse than ML when the kernel is poo…

View free PDFSource page
arxivcs.CVcs.AIcs.LGcs.MM2026-07-17

Debiasing Text-to-Image Evaluation via Implicit Cultural Alignment Reward Modeling

Bo-An Chang, Yu-Chih Chen

As Text-to-Image (T2I) systems rapidly advance, evaluating the cultural authenticity of synthesized content has become increasingly important for fair and trustworthy generative AI. Existing T2I evaluation metrics and multimodal judges often rely on visual-semantic representation…

View free PDFSource page
arxivcs.LGcs.AI2026-07-17

Agentic Calibration of Grey-Box Simulation Models: An LLM-Driven Alternative

David Gómez-Guillén, Mireia Diaz, Josep Lluis Arcos, Jesús Cerquides

Calibration of grey-box simulation models is a constrained optimization problem in which model evaluations are expensive, the parameter space can be high-dimensional, and the search must respect plausibility constraints. Although the simulation code is fully available to the anal…

View free PDFSource page
arxivcs.AIcs.LGstat.ML2026-07-10

ConfidenceBench: Evaluating Confidence Calibration in Large Language Models

Matthew ffrench-Constant, Daniel Yang, Xinmeng Huang, Sanyam Kapoor

Large language models (LLMs) are increasingly deployed in settings where fluent but incorrect answers can be costly. In these settings, accuracy alone is insufficient: models must also know when they are likely to be wrong. We present ConfidenceBench, a calibration benchmark that…

View free PDFSource page