From Isolated Scoring to Collaborative Ranking: A Comparison-Native Framework for LLM-Based Paper Evaluation

TL;DR

This paper introduces a comparison-native framework for evaluating scientific papers using LLMs, shifting from absolute scoring to pairwise ranking to improve robustness and generalization across datasets.

Contribution

The authors propose CNPE, a novel framework that integrates comparison into data construction and model training for more reliable paper evaluation.

Findings

Achieves 21.8% relative improvement over baseline DeepReview-14B.

Demonstrates robust generalization to five unseen datasets.

Utilizes a graph-based similarity ranking for informative pair sampling.

Abstract

Large language models (LLMs) are currently applied to scientific paper evaluation by assigning an absolute score to each paper independently. However, since score scales vary across conferences, time periods, and evaluation criteria, models trained on absolute scores are prone to fitting narrow, context-specific rules rather than developing robust scholarly judgment. To overcome this limitation, we propose shifting paper evaluation from isolated scoring to collaborative ranking. In particular, we design a $\textbf{C}$omparison-$\textbf{N}$ative framework for $\textbf{P}$aper $\textbf{E}$valuation ($\textbf{CNPE}$), integrating comparison into both data construction and model learning. We first propose a graph-based similarity ranking algorithm to facilitate the sampling of more informative and discriminative paper pairs from a collection. We then enhance relative quality judgment through supervised fine-tuning and reinforcement learning with comparison-based rewards. At inference, the model performs pairwise comparisons over sampled paper pairs and aggregates these preference signals into a global relative quality ranking. Experimental results demonstrate that our framework achieves an average relative improvement of 21.8% over the strong baseline DeepReview-14B, while exhibiting robust generalization to five previously unseen datasets. Our code is available at https://github.com/ECNU-Text-Computing/ComparisonReview.