LLM-as-a-Verifier: A General-Purpose Verification Framework | alphaXiv

Researchers from Stanford, UC Berkeley, and NVIDIA Research developed LLM-as-a-Verifier, a training-free framework that leverages fine-grained logit distributions for continuous feedback,...

Abstract Scaling pre-training, post-training, and test-time compute have become the central paradigms for improving the capabilities of LLMs. In this work, we identify verification, the ability to determine the correctness of a solution, as a new scaling axis. To unlock this and demonstrate its effectiveness, we introduce LLM-as-a-Verifier, a general-purpose verification framework that provides fine-grained feedback for agentic tasks without requiring additional training. Unlike standard LM judges that prompt LLMs to produce discrete scores for candidate solutions, LLM-as-a-Verifier computes t

saved by

• Ziran Yang

related reading

• Reasoning in General Domains without Verifiersarxiv.org
• State of RL for reasoning LLMs | A. Weersaweers.de
• Modifying LLM Beliefs with Synthetic Document Finetuningalignment.anthropic.com
• [2510.13786] The Art of Scaling Reinforcement Learning Compute for LLMsarxiv.org
• The Verification Horizon: No Silver Bullet for Coding Agent Rewardsarxiv.org
• Let's Verify Step by Steparxiv.org
• [2504.13837] Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?arxiv.org
• 2401.10020.pdfarxiv.org
• You Only Need Minimal RLVR Training: Extrapolating LLMs via Rank-1 Trajectoriesarxiv.org
• Evaluation-Conditioned Training: Teaching Models to Generalize to Stronger Oversight Regimesarxiv.org
• [2412.00543] Evaluating the Consistency of LLM Evaluatorsarxiv.org
• Large Language Models Must Be Taught to Know What They Don't Knowarxiv.org