InternReviewer & InternAdvocate: Objective Reward and Evaluation for Agentic Reinforcement Learning in Peer Review and Rebuttal

Researchers introduced InternReviewer and InternAdvocate, scholarly agents trained with an agentic reinforcement learning framework that uses a unified reward system to improve peer review and rebuttal generation. The framework, detailed in arXiv:2608.28612v1, incorporates a large-scale scholarly dataset and a high-efficiency arXiv retrieval tool, with multi-dimensional criteria including reference-anchored semantic alignment, structural compliance, and citation verification to eliminate hallucinations. Experimental results showed significant improvements in reasoning depth and citation accuracy.

arXiv:2608.28612v1 Announce Type: new Abstract: Generating professional scholarly content, such as peer reviews and rebuttals, requires an intricate synergy between domain reasoning and factual grounding. This work presents a comprehensive framework for the development and evaluation of specialized scholarly agents, InternReviewer and InternAdvocate. We first establish a large-scale, high-quality scholarly dataset and integrate a high-efficiency arXiv retrieval tool to enable active evidence gathering. To optimize these agents, we implement an agentic Reinforcement Learning RL paradigm driven by a unified objective metric and reward system. This system avoids the biases of subjective model-based judging by employing multi-dimensional criteria, including reference-anchored semantic alignment, structural compliance, and a strict verification mechanism that cross-checks citations against real-time interaction logs to eliminate hallucinations. Experimental results demonstrate that agents trained within this closed-loop framework exhibit significant improvements in reasoning depth and citation accuracy.