Graph-Guided Passage Retrieval for Author-Centric Structured Feedback
Obtaining high-quality, pre-submission feedback is a critical bottleneck in the academic publication lifecycle for researchers. We introduce AutoRev, an automated author-centric feedback system that generates structured, actionable guidance prior to formal peer review. AutoRev employs a graph-based retrieval-augmented generation framework that models each paper as a hierarchical document graph, integrating textual and structural representations to retrieve salient content efficiently. By leveraging graph-based passage retrieval, AutoRev substantially reduces LLM input context length, leading to higher-quality feedback generation. Experimental results demonstrate that AutoRev significantly outperforms baselines across multiple automatic evaluation metrics, while achieving strong performance in human evaluations. Code will be released upon acceptance.