ConceptPRM / Process Reward Model
Process Reward Model
A model that assigns reward signals to each individual intermediate step within a reasoning chain, representing an evolution of RLHF that refines rewards from final outcomes to each step of the process. It is particularly effective for tasks requiring rigorous logical chains, such as mathematics and programming.
Timeline (last 90 days)
Aug 23
OpenAI published research in 2023 showing that process reward models can significantly improve the reliability of mathematical reasoning compared to outcome supervision.
Unverified60%
All Facts (1)
Source Articles
How Test-Time Scaling Allocates Depth, Width, and Feedback in LLM ReasoningOct 12Why Does PRM-Pruned Fragment Grafting Fail? An Empirical Analysis Across Three Reasoning ModelsOct 2Deep Dive into GRPO: An Efficient Alignment Method for LLM Reinforcement LearningSep 23NeoHorse-1-4B Hands-On Review: The Local Small Model That Trains ItselfSep 16Gemini 3.5 Pro Rebuilt from Scratch: Delays, Talent Exodus, and the AI Competition LandscapeJul 21Interpretability: The Science and Practice of Opening the Neural Network Black BoxJul 19