ReviewGrounder: Improving Review Substantiveness with Rubric-Guided, Tool-Integrated Agents

TL;DR

ReviewGrounder enhances AI-generated peer reviews by integrating explicit rubrics and contextual grounding, leading to more substantive, evidence-based feedback that aligns better with human judgments.

Contribution

It introduces REVIEWBENCH for evaluating reviews and REVIEWGROUNDER, a multi-agent framework that improves review quality through rubric-guided drafting and grounding stages.

Findings

REVIEWGROUNDER outperforms baselines in review quality metrics.

The framework shows improved alignment with human judgments.

Using larger models further enhances review quality.

Abstract

The rapid rise in AI conference submissions has driven increasing exploration of large language models (LLMs) for peer review support. However, LLM-based reviewers often generate superficial, formulaic comments lacking substantive, evidence-grounded feedback. We attribute this to the underutilization of two key components of human reviewing: explicit rubrics and contextual grounding in existing work. To address this, we introduce REVIEWBENCH, a benchmark evaluating review text according to paper-specific rubrics derived from official guidelines, the paper's content, and human-written reviews. We further propose REVIEWGROUNDER, a rubric-guided, tool-integrated multi-agent framework that decomposes reviewing into drafting and grounding stages, enriching shallow drafts via targeted evidence consolidation. Experiments on REVIEWBENCH show that REVIEWGROUNDER, using a Phi-4-14B-based drafter and a GPT-OSS-120B-based grounding stage, consistently outperforms baselines with substantially stronger/larger backbones (e.g., GPT-4.1 and DeepSeek-R1-670B) in both alignment with human judgments and rubric-based review quality across 8 dimensions. The code is available \href{https://github.com/EigenTom/ReviewGrounder}{here}.