PRM-800K

The PRM-800K dataset is a benchmark that contains examples designed to evaluate large language models on their ability to assess causal and coverage faithfulness in chain-of-thought reasoning.

Papers using PRM-800K (6)

C2-Faith: Benchmarking LLM Judges for Causal and Coverage Faithfulness in Chain-of-Thought Reasoning2026Process Reward Models That Think2025SPC: Evolving Self-Play Critic via Adversarial Games for LLM Reasoning2025Error Typing for Smarter Rewards: Improving Process Reward Models with Error-Aware Hierarchical Supervision2025DeepCritic: Deliberate Critique with Large Language Models2025Understanding Chain-of-Thought in LLMs through Information Theory2024 · 1 cites