Star on GitHubStar3.2k stars

2026 · Conference paper

Rethinking Neural Decompilation: Candidate Selection as the Hidden Bottleneck

The paper compares methods for selecting LLM4Decompile's reconstructed code candidates and evaluates fine-tuning on an OLLVM-obfuscated binary dataset.

Scope and limitations

The evaluation covers LLM4Decompile on HumanEval-Decompile, ExeBench, and the ExeBench-Obfuscated dataset.

Why it fits this collection

It evaluates neural decompilation, candidate selection, and code reconstruction from obfuscated binaries.

Cite

@inproceedings{paper-rethinking-neural-decompilation,
  title = {{Rethinking Neural Decompilation: Candidate Selection as the Hidden Bottleneck}},
  author = {{Felix Mächtle} and {Gautham Sriram} and {Manjil Neupane} and {Nils Loose} and {Thomas Eisenbarth}},
  year = {2026},
  url = {https://doi.org/10.1007/978-981-92-6837-5_38},
  booktitle = {ICICS 2026},
  doi = {10.1007/978-981-92-6837-5_38},
  eprint = {2611.02788},
  archivePrefix = {arXiv},
  month = oct
}

Read in a terminal: hf papers read 2611.02788

Publication and review details

Venue
ICICS 2026
Status
Conference paper
Publication / posting date
October 23, 2026
Date source
Publication date in the Crossref record for the DOI.
First arXiv submission
November 2, 2026
arXiv
2611.02788
Reviewed arXiv version
v1
DOI
10.1007/978-981-92-6837-5_38

The venue and date come from the Crossref record for the DOI; the PDF is the arXiv version.

Sources reviewed

Reviewed against the paper and its primary sources. Cite the original authors using the linked paper or BibTeX.

Report an outdated entry Suggest a translation fix

Primary sources

  1. https://arxiv.org/abs/2611.02788
  2. https://doi.org/10.1007/978-981-92-6837-5_38
  3. https://api.crossref.org/works/10.1007/978-981-92-6837-5_38
↑↓ move↵ open⌘↵ new tab