Episode
Process Rewards with Learned Reliability
- Podcast
- Daily Paper Cast
- Published
- May 21, 2026
- Duration seconds
- 1401
- Processing state
not_requested- Canonical source
- https://share.transistor.fm/s/dfadd2c7
Actions
POST https://stenobird.com/v1/public/podcasts/daily-paper-cast-7079649/episodes/process-rewards-with-learned-reliability/transcription-requests
Idempotently request low-priority transcript generation for this episode.GET https://stenobird.com/podcast/daily-paper-cast-7079649/process-rewards-with-learned-reliability.md
Read the agent-friendly Markdown representation of this episode resource.
Summary
🤗 Upvotes: 46 | cs.CL, cs.AI, cs.LG Authors: Jinyuan Li, Langlin Huang, Chengsong Huang, Shaoyang Xu, Donghong Cai, Yuyi Yang, Wenxuan Zhang, Jiaxin Huang Title: Process Rewards with Learned Reliability Arxiv: http://arxiv.org/abs/2605.15529v1 Abstract: Process Reward Models (PRMs) provide step-level feedback for reasoning, but current PRMs usually output only a single reward score for each step. Downstream methods must therefore treat imperfect step-level reward predictions as reliable decision signals, with no indication of when these predictions should be trusted. We propose BetaPRM, a distributional PRM that predicts both a step-level success probability and the reliability of that prediction. Given step-success supervision from Monte Carlo continuations, BetaPRM learns a Beta belief that explains the observed number of successful continuations through a Beta-Binomial likelihood, rather than regressing to the finite-sample success ratio as a point target. This learned reliability signal indicates when a step reward should be trusted, enabling downstream applications to distinguish reliable rewards from uncertain ones. As one application, we introduce Adaptive Computation Allocation (ACA) for PRM-guided Best-of-N reasoning. ACA uses the learned reliability signal to stop when a high-reward solution is reliable and to spend additional computation on uncertain candidate prefixes. Experiments across four backbones and four reasoning benchmarks show that BetaPRM improves PRM-guided Best-of-N selection while preserving standard step-level error detection. Built on this signal, ACA improves the accuracy--token tradeoff over fixed-budget Best-of-16, reducing token usage by up to 33.57% while improving final-answer accuracy.