Rethinking Neural Decompilation: Candidate Selection as the Hidden Bottleneck

Summary

This paper studies why neural decompilers based on LLM4Decompile fall short of their best-sample accuracy on obfuscated binaries. It shows correct reconstructions are usually already among the model's samples, reaching 93% pass@100 on HumanEval-Decompile. A generate-then-select pipeline using negative perplexity and execution-guided selection, plus one epoch of fine-tuning on the released ExeBench-Obfuscated dataset (362,647 OLLVM samples), raises pass@1 on obfuscated binaries to 88.8%.