When Models Edit Too Much: On the Fidelity of Minimal Code Edits

arXiv:2609.04061 · cs.SE, cs.AI, cs.CL · Submitted 2026-09-03 · Read on arXiv

cs.SE, cs.AI, cs.CL

Submitted: 2026-09-03

Updated: 2026-09-03

Code: https://github.com/nreHieW/over-editing

Terminology

• Fidelity — 27× in this paper · explained in 2 episodes
• Benchmark — 22× in this paper · explained in 2 episodes
• Ablation — 11× in this paper · explained in 1 episode
• Fine-Tuning — 7× in this paper · explained in 5 episodes
• Generalization — 6× in this paper · explained in 13 episodes
• Reinforcement Learning — 6× in this paper · explained in 4 episodes
• Robustness — 3× in this paper · explained in 5 episodes
• Aggregation — 2× in this paper · explained in 1 episode
• Benchmarking — 2× in this paper · explained in 3 episodes
• Direct Preference Optimization — 2× in this paper
• GRPO — 2× in this paper · explained in 6 episodes
• Trace — 2× in this paper · explained in 2 episodes
• Adaptation — 1× in this paper · explained in 1 episode
• Agentic Coding — 1× in this paper · explained in 1 episode
• Decoupling — 1× in this paper · explained in 2 episodes
• Direct Preference Optimization (DPO) — 1× in this paper · explained in 9 episodes
• Foundation Model — 1× in this paper · explained in 3 episodes
• Foundation Models — 1× in this paper · explained in 7 episodes
• Granularity Mismatch — 1× in this paper · explained in 2 episodes
• Large Language Models (LLMs) — 1× in this paper · explained in 6 episodes
• LoRA Adapters — 1× in this paper · explained in 3 episodes
• Online Reinforcement Learning — 1× in this paper · explained in 2 episodes
• Positional Bias — 1× in this paper · explained in 2 episodes
• Tokenization — 1× in this paper · explained in 2 episodes
• Verification — 1× in this paper · explained in 2 episodes

Sources

• Program Synthesis with Large Language Models
• Can It Edit? Evaluating the Ability of Large Language Models to Follow Code Editing Instructions
• Evaluating Large Language Models Trained on Code
• EDIT-Bench: Evaluating LLM Abilities to Perform Real-World Instructed Code Edits
• SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training
• Less is More: Adaptive Program Repair with Bug Localization and Preference Learning
• DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models
• DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
• DeepSeek-V3 Technical Report
• Edit, But Verify: An Empirical Audit of Instructed Code-Editing Benchmarks
• CoreCodeBench: Decoupling Code Intelligence via Fine-Grained Repository-Level Tasks
• Scaling Reasoning, Losing Control: Evaluating Instruction Following in Large Reasoning Models
• GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models
• CodeEditorBench: Evaluating Code Editing Capability of Large Language Models
• SWE-Perf: Can Language Models Optimize Code Performance on Real-World Repositories?
• Qwen2.5-Coder Technical Report
• LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code
• SWE-bench: Can Language Models Resolve Real-World GitHub Issues?
• QiMeng-PRepair: Precise Code Repair via Edit-Aware Reward Optimization
• TACO: Topics in Algorithmic COde generation dataset

Related papers

• Falsification-Based Verification of LLM-Generated Optimization Models: Sound Test Batteries and Their Detection Limits
• GitSkills: A Dataset of Agent Skills on GitHub
• SABER: Benchmarking Operational Safety of LLM Coding Agents in Stateful Project Workspaces
• PackMonitor: Enabling Zero Package Hallucinations Through Decoding-Time Monitoring
• IntentCoding: Amplifying User Intent in Code Generation
• Incentives and Outcomes in Bug Bounties