Learning Project-wise Subsequent Code Edits via Interleaving Neural-based Induction and Tool-based Deduction

2026-04-13

AI Summary

Existing large language models struggle to simultaneously achieve accuracy, cross-file awareness, and efficiency in project-level code editing: localized approaches are precise but lack global context, while global methods incur substantial computational overhead. This work proposes TRACE, a novel framework that synergistically integrates neural models with IDE tools by alternately performing semantic edits—predicted by a neural network—and syntactic edits—derived via refactoring and static analysis tools. TRACE introduces a fine-grained edit representation and a dynamic scheduling mechanism that judiciously determines when to invoke each tool. Experimental results demonstrate that TRACE significantly improves cross-file editing accuracy while maintaining high efficiency, outperforming state-of-the-art methods such as Cursor and CoEdPilot.

Technology Category

• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Machine Learning: Large Multimodal Models (LMMs)
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Large language models for search
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

In industrial and open-source software engineering tasks, developers often perform project-wise code editing tasks, including feature enhancement, refactoring, and bug fixing, where the leading AI models are expected to support the productivity. Hence, researchers and practitioners have proposed and adopted many LLM-based solutions to facilitate their real-world development. However, they largely suffer from the balance among predicting scope, accuracy, and efficiency. For example, solutions like Cursor achieve high accuracy only in a local editing scope while its performance drops on cross-file edits. In contrast, solutions like CoEdPilot exhibit efficiency limitations when used to predict project-wise edits. In this work, we propose TRACE (Tool-integrated RecommendAtion for Code Editing), a novel subsequent code editing solution to push the boundary of scope, accuracy, and efficiency. Our rationale lies in that code edits are triggered for either semantic or syntactic reasons. Therefore, TRACE predicts subsequent edits by interleaving neural-based induction for semantic edit prediction and tool-based deduction for syntactic edit prediction. The tools can be any IDE facilities, such as refactoring tools (e.g., rename) or linting tools (e.g., use-def), providing decent performance of deducing edit-location and edit-generation. Technically, we address the challenge of (1) when to interleave between neural-based and tool-based prediction and (2) how to further improve the performance of neural-based prediction. As for the former, we learn a neural model to detect when to invoke IDE editing tools. As for the latter, we propose a novel and fine-grained editing representation to further boost the performance of neural editing models. ......

Problem

• code editing
• project-wise edits
• large language models
• accuracy-efficiency trade-off
• cross-file editing

Innovation

• neural-symbolic integration
• code editing prediction
• IDE tool integration
• fine-grained edit representation
• project-wise code changes