# Megadiff: A Dataset of 600k Java Source Code Changes Categorized by Diff Size

arXiv: https://arxiv.org/abs/2108.04631 · HTML: https://mlsanity.com/paper/2108.04631/

## Key facts

- Cited by 10 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Aug 2021.

## Built on by (most cited first, 10 of 10)

- [OctoPack: Instruction Tuning Code Large Language Models](https://mlsanity.com/paper/2308.07124/) (2023)
- [Unifying the Perspectives of NLP and Software Engineering: A Survey on Language Models for Code](https://mlsanity.com/paper/2311.07989/) (2023)
- [RepairLLaMA: Efficient Representations and Fine-Tuned Adapters for Program Repair](https://mlsanity.com/paper/2312.15698/) (2023)
- [From Code Foundation Models to Agents and Applications: A Comprehensive Survey and Practical Guide to Code Intelligence](https://mlsanity.com/paper/2511.18538/) (2025)
- [RunBugRun -- An Executable Dataset for Automated Program Repair](https://mlsanity.com/paper/2304.01102/) (2023)
- [The Art of Repair: Optimizing Iterative Program Repair with Instruction-Tuned Models](https://mlsanity.com/paper/2505.02931/) (2025)
- [Out of Context: How important is Local Context in Neural Program Repair?](https://mlsanity.com/paper/2312.04986/) (2023)
- [Memory-Efficient Large Language Models for Program Repair with Semantic-Guided Patch Generation](https://mlsanity.com/paper/2410.16655/) (2024)
- [CodeLL: A Lifelong Learning Dataset to Support the Co-Evolution of Data and Language Models of Code](https://mlsanity.com/paper/2312.12492/) (2023)
- [On the Impact of Code Comments for Automated Bug-Fixing: An Empirical Study](https://mlsanity.com/paper/2601.23059/) (2026)

## Abstract (arXiv)

This paper presents Megadiff, a dataset of source code diffs. It focuses on Java, with strict inclusion criteria based on commit message and diff size. Megadiff contains 663 029 Java diffs that can be used for research on commit comprehension, fault localization, automated program repair, and machine learning on code changes.