Research
DecompileBench (Findings of ACL 2025)
A benchmark combining functions from real programs, runtime-aware validation, and LLM-assisted assessment of decompiled-code usability. Helps separate functional correctness from perceived understandability when comparing decompilers.
Requirements and limitations
- Findings of ACL 2025 paper. LLM-as-judge assessments and runtime correctness measure different properties; published results are not independent reproductions by this collection.
About this entry
- Type
- Research
- Category
- Learning resources
- Relationship
- Learning resource
- Targets
- Binary
Sources reviewed
The description was reviewed against primary sources. This does not establish runtime testing, security auditing or verified compatibility.