Decaf: Improving Neural Decompilation with Automatic Feedback and Search

• URL: http://arxiv.org/abs/2605.11501v1
• Date: Tue, 12 May 2026 04:21:26 GMT
• Title: Decaf: Improving Neural Decompilation with Automatic Feedback and Search
• Abstract summary: High-level syntax, identifiers, and custom data types are generally lost as the compiler translates human-readable code to low-level machine code.<n>Generative AI models are a natural fit for reconstructing high-level syntax, identifiers, and types.<n>We argue that compiler feedback can be used to dramatically improve the semantic correctness of neural decompiler outputs via search.
• Score: 64.02615739798277
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Decompilers are useful tools used in reverse engineering to understand compiled source code. Reconstructing source code from compiled binaries is a challenging task, because high-level syntax, identifiers, and custom data types are generally lost as the compiler translates human-readable code to low-level machine code. Deterministic decompilers are useful tools for binary analysis, but can struggle to infer idiomatic syntax and identifier names. Generative AI models are a natural fit for reconstructing high-level syntax, identifiers, and types, but they can still suffer by hallucinating improper programming constructs and semantics. Instead of attempting to improve neural decompilers with more data and more training, we argue that compiler feedback can be used to dramatically improve the semantic correctness of neural decompiler outputs via search. Our system, Decaf (DECompilation with Automated Feedback), raises the neural decompilation rate from 26.0% on ExeBench to 83.9% on the Real -O2 split without sacrificing similarity to the original source code. We also find our automatic feedback methodology is highly effective for improving weaker neural decompilation models.

Related papers

• Context-Guided Decompilation: A Step Towards Re-executability [50.71992919223209]
  Binary decompilation plays an important role in software security analysis, reverse engineering and malware understanding.<n>Recent advances in large language models (LLMs) have enabled neural decompilation, but the generated code is typically only semantically plausible.<n>We propose ICL4Decomp, a hybrid decompilation framework that leverages in-context learning (ICL) to guide LLMs toward generating re-executable source code.
  arXiv  Detail & Related papers  (2025-11-03T17:21:39Z)
• ReF Decompile: Relabeling and Function Call Enhanced Decompile [50.86228893636785]
  The goal of decompilation is to convert compiled low-level code (e.g., assembly code) back into high-level programming languages.<n>This task supports various reverse engineering applications, such as vulnerability identification, malware analysis, and legacy software migration.
  arXiv  Detail & Related papers  (2025-02-17T12:38:57Z)
• Idioms: Neural Decompilation With Joint Code and Type Definition Prediction [7.421408987075001]
  We introduce a new dataset, Realtype, that includes substantially more complicated and realistic types than existing neural decompilation benchmarks.<n>We show that our approach yields state-of-the-art results in neural decompilation.
  arXiv  Detail & Related papers  (2025-02-06T22:13:40Z)
• Fast, Fine-Grained Equivalence Checking for Neural Decompilers [7.421408987075001]
  We introduce codealign, a novel instruction-level code equivalence technique designed for neural decompilers.<n>We show how codealign generates equivalence alignments, then evaluate codealign by comparing it with symbolic execution.
  arXiv  Detail & Related papers  (2025-01-08T19:59:48Z)
• LILO: Learning Interpretable Libraries by Compressing and Documenting Code [71.55208585024198]
  We introduce LILO, a neurosymbolic framework that iteratively synthesizes, compresses, and documents code. LILO combines LLM-guided program synthesis with recent algorithmic advances in automated from Stitch. We find that AutoDoc boosts performance by helping LILO's synthesizer to interpret and deploy learned abstractions.
  arXiv  Detail & Related papers  (2023-10-30T17:55:02Z)
• Guess & Sketch: Language Model Guided Transpilation [59.02147255276078]
  Learned transpilation offers an alternative to manual re-writing and engineering efforts. Probabilistic neural language models (LMs) produce plausible outputs for every input, but do so at the cost of guaranteed correctness. Guess & Sketch extracts alignment and confidence information from features of the LM then passes it to a symbolic solver to resolve semantic equivalence.
  arXiv  Detail & Related papers  (2023-09-25T15:42:18Z)
• Boosting Neural Networks to Decompile Optimized Binaries [13.255618541522436]
  Decompilation aims to transform a low-level program language (LPL) into its functionally-equivalent high-level program language (HPL) We propose a novel learning-based approach named NeurDP, that targets compiler-optimized binaries.
  arXiv  Detail & Related papers  (2023-01-03T06:45:54Z)
• Beyond the C: Retargetable Decompilation using Neural Machine Translation [5.734661402742406]
  We develop a prototype decompiler that is easily retargetable to new languages. We examine the impact of parameters such as tokenization and training data selection on the quality of decompilation. We will release our training data, trained decompilation models, and code to help encourage future research into language-agnostic decompilation.
  arXiv  Detail & Related papers  (2022-12-17T20:45:59Z)
• Improving type information inferred by decompilers with supervised machine learning [0.0]
  In software reverse engineering, decompilation is the process of recovering source code from binary files. We build different classification models capable of inferring the high-level type returned by functions. Our system is able to predict function return types with a 79.1% F1-measure, whereas the best decompiler obtains a 30% F1-measure.
  arXiv  Detail & Related papers  (2021-01-19T11:45:46Z)
• Contrastive Code Representation Learning [95.86686147053958]
  We show that the popular reconstruction-based BERT model is sensitive to source code edits, even when the edits preserve semantics. We propose ContraCode: a contrastive pre-training task that learns code functionality, not form.
  arXiv  Detail & Related papers  (2020-07-09T17:59:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.