ReFeed: Multi-dimensional Summarization Refinement with Reflective Reasoning on Feedback
Taewon Yun, Jihwan Oh, Hyangsuk Min, Yuho Lee, Jihwan Bang, Jason Cai, Hwanjun Song
2025-03-31
Summary
This paper is about improving how AI summarizes things by having it think about the feedback it gets and make adjustments.
What's the problem?
It's hard for AI to summarize things well across different areas, and giving feedback in one area can mess up the summary in another.
What's the solution?
The researchers developed a system that helps AI think about feedback and make better summaries by considering multiple areas at once.
Why it matters?
This work matters because it can lead to AI that can create more accurate and comprehensive summaries of complex information.
Abstract
Summarization refinement faces challenges when extending to multi-dimension. In this paper, we introduce ReFeed, a powerful summarization refinement pipeline that enhances multiple dimensions through reflective reasoning on feedback. To achieve this, we release SumFeed-CoT, a large-scale Long-CoT-based dataset optimized for training a lightweight model with reflective reasoning. Our experiments reveal how the number of dimensions, feedback exposure, and reasoning policy influence refinement performance, highlighting reflective reasoning and simultaneously addressing multiple feedback is crucial to mitigate trade-off between dimensions. Furthermore, ReFeed is robust to noisy feedback and feedback order. Lastly, our finding emphasizes that creating data with a proper goal and guideline constitutes a fundamental pillar of effective reasoning. The dataset and model will be released.