PRISM is a contrastive, policy-aware training framework for process reward models that reduces false positives by 22% on PRMBench and boosts downstream accuracy up to 33% in Best-of-N selection by learning reliable relative comparisons instead of pointwise labels.
hub
arXiv preprint arXiv:2504.00891 , year=
13 Pith papers cite this work. Polarity classification is still indexing.
13
Pith papers citing it
hub tools
citation-role summary
background 2
citation-polarity summary
roles
background 2polarities
background 2representative citing papers
Temporal scheduling of credit allocation criteria over RLVR training, using trajectory percentiles to target heterogeneous behaviors, yields more stable policy entropy and better reasoning benchmark results than static allocation.
Unsupervised PRMs derived from LLM probabilities achieve up to 15% better error detection than LLM judges and match supervised PRMs in verification and RL tasks.
DataPRM is an environment-aware generative process reward model that improves LLM data analysis agents by 7-11% on benchmarks via active verification and reflection-aware ternary rewards.
This survey introduces the Generate-Filter-Control-Replay (GFCR) taxonomy to structure rollout pipelines for RL-based post-training of reasoning LLMs.
OpenClaw-RL recovers evaluative and directive signals from next-state interactions to enable online RL training of agents across terminal, GUI, SWE, and tool environments via a server-client architecture and hybrid objective.
PROF curates RL training data via PRM-ORM consistency to improve both final-answer accuracy and intermediate reasoning quality while reducing reliance on strong process reward models.
Archer introduces response-level entropy normalization and differentiated clipping/KL regularization in RLVR to encourage exploration on reasoning tokens while stabilizing knowledge tokens, yielding gains in pass@1 and pass@K on reasoning benchmarks.
CoLD mitigates length bias in process reward models for mathematical reasoning via counterfactual guidance, length penalties, bias estimation, and joint training, improving step selection accuracy and conciseness on MATH500 and GSM-Plus while boosting downstream RL performance.
A survey of test-time scaling for multimodal foundation models that introduces a three-way taxonomy of sampling, feedback, and search approaches along with applications and benchmarks.
TrOPD stabilizes on-policy distillation for LLMs with trust-region learning, outlier estimation, and off-policy guidance, outperforming prior OPD methods on reasoning and code benchmarks.
citing papers explorer
-
The Hidden Bias of Process Reward Models:PRISM for Rewarding the Right Reasoning
PRISM is a contrastive, policy-aware training framework for process reward models that reduces false positives by 22% on PRMBench and boosts downstream accuracy up to 33% in Best-of-N selection by learning reliable relative comparisons instead of pointwise labels.
-
Not only where, But when: Temporal Scheduling for RLVR
Temporal scheduling of credit allocation criteria over RLVR training, using trajectory percentiles to target heterogeneous behaviors, yields more stable policy entropy and better reasoning benchmark results than static allocation.
-
Unsupervised Process Reward Models
Unsupervised PRMs derived from LLM probabilities achieve up to 15% better error detection than LLM judges and match supervised PRMs in verification and RL tasks.
-
Generate, Filter, Control, Replay: A Comprehensive Survey of Rollout Strategies for LLM Reinforcement Learning
This survey introduces the Generate-Filter-Control-Replay (GFCR) taxonomy to structure rollout pipelines for RL-based post-training of reasoning LLMs.
-
Trust Region On-Policy Distillation
TrOPD stabilizes on-policy distillation for LLMs with trust-region learning, outlier estimation, and off-policy guidance, outperforming prior OPD methods on reasoning and code benchmarks.