2605.29886v1 May 28, 2026 cs.CL

CRITIC-R1: Learning Structured Critics for Retrieval-Augmented Generation

Xingcheng Fu
Xingcheng Fu
Guangxi Normal University
Citations: 921
h-index: 15
Jianxin Li
Jianxin Li
Citations: 326
h-index: 10
Wen Xiao
Wen Xiao
Citations: 151
h-index: 6
Chuanyue Yu
Chuanyue Yu
Citations: 12
h-index: 2
Qingyu Sun
Qingyu Sun
Citations: 39
h-index: 3
Runhua Xu
Runhua Xu
Citations: 12
h-index: 2
Ziwei Zhang
Ziwei Zhang
Citations: 1,726
h-index: 23

Retrieval-augmented generation (RAG) improves knowledge-intensive question answering by incorporating external evidence. However, existing RAG methods still suffer from hallucinations and subtle reasoning errors. Recent studies introduce external critics to refine RAG outputs, yet they often provide coarse-grained and weakly structured feedback, exhibit over-aggressive intervention, and lead to noisy and unreliable refinement, limiting their effectiveness for correction. To tackle these issues, we propose CRITIC-R1, a structured critic framework that formulates and learns RAG critique as an explicit error diagnosis problem using reinforcement learning (RL). Our framework categorizes common RAG errors into multiple diagnostic dimensions, including verdict, error location, reasoning analysis, and fix generation. To learn these capabilities, we design two reward functions: Conservative Judgement Alignment (CJA) first encourages calibrated high-level judgements while mitigating the over-aggressive phenomenon, whereas Diagnostic Quality Alignment (DQA) further improves fine-grained diagnostic feedback through gated rewards. We train the critic model using GRPO-based RL with process-level supervision collected from external LLM teacher models. Experiments across five QA benchmarks show that CRITIC-R1 consistently improves answer quality over strong RAG baselines. Our source code is available at https://anonymous.4open.science/r/critic-r1-FCB0

0 Citations
0 Influential
11.5 Altmetric
57.5 Score
Original PDF

No Analysis Report Yet

This paper hasn't been analyzed by Gemini yet.

Log in to request an AI analysis.

댓글

댓글을 작성하려면 로그인하세요.

아직 댓글이 없습니다. 첫 번째 댓글을 남겨보세요!