XtraGPT: Context-Aware and Controllable Academic Paper Revision
Nuo Chen, Andre Lin HuiKai, Jiaying Wu, Junyi Hou, Zining Zhang, Qian Wang, Xidong Wang, Bingsheng He
TL;DR
The paper tackles the challenge of making LLMs effective for iterative, high-quality academic paper revision by introducing a context-aware, criteria-guided human–AI collaboration framework. It presents ReviseQA, a 7,000-paper dataset with 140,800 instruction–revision pairs, and XtraGPT, an open-source family of models (1.5B–14B) trained to perform targeted, section-level revisions while preserving global coherence. Through CPT-based training and LC-Win Rate evaluation with LLM-as-a-Judge, the work demonstrates that XtraGPT outperforms same-scale baselines and approaches proprietary systems, with human evaluators confirming rationale-aligned improvements and higher adoption likelihood. The framework emphasizes controllability, iterative refinement, and full-document context, highlighting practical impact for researchers seeking reliable, transparent AI-assisted paper revision while acknowledging limitations and ethical considerations for deployment.
Abstract
Despite the growing adoption of large language models (LLMs) in academic workflows, their capabilities remain limited to support high-quality scientific writing. Most existing systems are designed for general-purpose scientific text generation and fail to meet the sophisticated demands of research communication beyond surface-level polishing, such as conceptual coherence across sections. Furthermore, academic writing is inherently iterative and revision-driven, a process not well supported by direct prompting-based paradigms. To address these scenarios, we propose a human-AI collaboration framework for academic paper revision centered on criteria-guided intent alignment and context-aware modeling. To validate the framework, we curate a dataset of 7,000 research papers from top-tier venues annotated with 140,000 instruction-response pairs that reflect realistic, section-level scientific revisions. We instantiate the framework in XtraGPT, the first suite of open-source LLMs (1.5B to 14B parameters) for context-aware, instruction-guided writing assistance. Extensive experiments validate that XtraGPT significantly outperforms same-scale baselines and approaches the quality of proprietary systems. Both automated preference assessments and human evaluations confirm the effectiveness of XtraGPT in improving scientific drafts.