Decompile-Eval

This is the evaluation benchmark of LLM4Decompile project. About Decompile-Bench-Eval includes manually crafted binaries from the well-established HumanEval and MBPP, alongside the compiled GitHub repositories released after 2025 to mitigate data leakage issues. C and C++ Support: These datasets include both C and C++

Papers using Decompile-Eval (2)

SALT4Decompile: Inferring Source-level Abstract Logic Tree for LLM-Based Binary Decompilation2025Self-Constructed Context Decompilation with Fined-grained Alignment Enhancement2024 · 1 cites