# Augmenting Decompiler Output with Learned Variable Names and Types

> Introduces DIRTY, a Transformer model that recommends variable names and types from decompiled code and layout information, together with the DIRT corpus for training and evaluation.

- Authors: Qibin Chen; Jeremy Lacomis; Edward J. Schwartz; Claire Le Goues; Graham Neubig; Bogdan Vasilescu
- Venue: USENIX Security 2022
- Status: Conference paper
- Publication / posting date: August 2022
- Date source: USENIX proceedings BibTeX publication month
- First arXiv submission: August 13, 2021
- arXiv: 2108.06363
- Reviewed arXiv version: v1
- Sources reviewed: 2026-10-16
- Added to the catalog: 2026-10-16
- Original paper: <https://www.usenix.org/conference/usenixsecurity22/presentation/chen-qibin>
- PDF: <https://arxiv.org/pdf/2108.06363>
- Code: <https://github.com/CMUSTRUDEL/DIRTY>
- Project: <https://dirtdirty.github.io/explorer.html>
- BibTeX: <https://awesomerea.org/papers/citations/paper-dirty.bib>
- Catalog page: <https://awesomerea.org/papers/paper-dirty/>
- Record JSON: <https://awesomerea.org/papers/paper-dirty/record.json>
- Chinese version: [简体中文](https://awesomerea.org/zh/papers/paper-dirty/index.md)

## Scope and limitations

The original pipeline is built around C programs and Hex-Rays output. Predicted names and types are learned hypotheses, not guaranteed original symbols. The published paper is from 2022, although its arXiv preprint first appeared in 2021.

## Why it fits this collection

Extends name recovery to types, making it directly relevant to assessing semantic annotations added by binary-analysis assistants.

## Publication and review notes

Year and publication month come from the final USENIX conference record; the arXiv date is not substituted for publication.

## Primary sources

1. <https://www.usenix.org/conference/usenixsecurity22/presentation/chen-qibin>
2. <https://arxiv.org/abs/2108.06363>
3. <https://github.com/CMUSTRUDEL/DIRTY>
4. <https://github.com/CMUSTRUDEL/DIRTY/blob/5f8e9494d52ef8788f0834342097f36462eeb2eb/README.md>
5. <https://edmcman.github.io/papers/usenix22.pdf>

## About this entry

Reviewed against the paper and its primary sources. Cite the original authors using the linked paper or BibTeX.
Original editorial content: MIT. Linked projects, papers and assets retain their own rights and licenses.