Newest papers

ExeBench-FHD: Addressing Fidelity Issues in Decompiler Evaluation

Akarsh Ashok, Dinil Mon Divakaran, Vivek Balachandran · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.28212 · cited by 0 works (OpenAlex) · open-access PDF

Russian

ExeBench является одним из наиболее широко используемых эталонов данных для оценки декомпиляторов на основе LLM. Мы анализируем ExeBench и идентифицируем девять проблем, которые влияют на то, насколько верно тест измеряет декомпилятор. Три из них мешают LLM-декомпиляторам восстанавливать код C из функции ExeBench. Это двусмысленный порядок параметров на оптимизаторе, когда смешиваются типы целых и плавающих точек, неиспользуемые параметры, которые могут падать в оптимизированной сборке, и константы, хранящиеся за пределами кода сборки функции. Еще шесть влияют на валидность теста: непустые функции без возврата, типы держателя места, статические или инлайн

Machine translation (Meta M2M100); the English original follows.

Abstract

ExeBench is one of the most widely used benchmarks for evaluating LLM-based decompilers. We analyze ExeBench and identify nine issues that affect how faithfully a test measures a decompiler. Three of them prevent LLM decompilers from recovering C code from ExeBench's function. These are ambiguous parameter order when integer and floating-point types are mixed, unused parameters that optimized assembly may drop, and constants stored outside the function's assembly code. Six more affect test validity: non-void functions without a return, placeholder types, static or inline functions removed by the optimizer, functions named main, trivial bodies and truncated definitions. We present ExeBench-FHD, a backward-compatible revision of ExeBench that adds 16 new fields: call-site assembly, data sections, test wrappers, linked-binary disassembly and five warning flags. In ExeBench, 38.9% of functions have at least one of the five flags; dropping records with any of the four warning flags (all except unused_param, which marks functions that the new wrappers make evaluable) keeps 63.9% of the corpus. In an evaluation at -O2 on four architectures, the new fields raise GPT 6 Sol's re-execution rate by up to 26.5%, 19.1% and 21.0% on functions that read data sections, mix parameter types and have unused parameters, respectively.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Computer science, Fidelity, Code (set theory), Test (biology), Integer (computer science), Order (exchange)

Is this paper real? This page was built from OpenAlex and arXiv records, not generated. Software can verify any reference in one call: GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.28212, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txt

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange