Pith. sign in

REVIEW 12 cited by

LLM4Decompile: Decompiling Binary Code with Large Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2403.05286 v3 pith:GNCYCXSP submitted 2024-03-08 cs.PL cs.CL

LLM4Decompile: Decompiling Binary Code with Large Language Models

classification cs.PL cs.CL
keywords codemodelsbinaryllm4decompileghidradecompilationdecompilelanguage
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Decompilation aims to convert binary code to high-level source code, but traditional tools like Ghidra often produce results that are difficult to read and execute. Motivated by the advancements in Large Language Models (LLMs), we propose LLM4Decompile, the first and largest open-source LLM series (1.3B to 33B) trained to decompile binary code. We optimize the LLM training process and introduce the LLM4Decompile-End models to decompile binary directly. The resulting models significantly outperform GPT-4o and Ghidra on the HumanEval and ExeBench benchmarks by over 100% in terms of re-executability rate. Additionally, we improve the standard refinement approach to fine-tune the LLM4Decompile-Ref models, enabling them to effectively refine the decompiled code from Ghidra and achieve a further 16.2% improvement over the LLM4Decompile-End. LLM4Decompile demonstrates the potential of LLMs to revolutionize binary code decompilation, delivering remarkable improvements in readability and executability while complementing conventional tools for optimal results. Our code, dataset, and models are released at https://github.com/albertan017/LLM4Decompile

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 12 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Checked Program Recovery from Execution Video: A Sound Oracle for Untrusted Generators

    cs.SE 2026-07 conditional novelty 7.0

    Vid2Prog recovers Scratch programs from execution videos via a sound oracle that certifies lens-equivalence with zero false accepts on 246 test pairs and 80% certificate rate for in-vocabulary cases while abstaining o...

  2. Agentic Vulnerability Reasoning on COTS Binaries

    cs.CR 2026-05 accept novelty 7.0

    SLYP agentic pipeline discovers race condition vulnerabilities in Windows COM binaries and generates debugger-verified PoCs, scoring 0.973 F1 on a 40-case benchmark and finding 28 new confirmed vulnerabilities in prod...

  3. Constraint-Guided Multi-Agent Decompilation for Executable Binary Recovery

    cs.SE 2026-04 unverdicted novelty 7.0

    A constraint-guided multi-agent system turns raw decompiler output into re-executable code at 84-97% success rates, outperforming prior LLM decompilation methods on real binaries.

  4. Can LLMs Deobfuscate Binary Code? A Systematic Analysis of Large Language Models into Pseudocode Deobfuscation

    cs.SE 2026-04 unverdicted novelty 7.0

    LLM deobfuscation of binaries to pseudocode depends more on reasoning ability and task-specific fine-tuning than on model size, with reasoning models showing robustness across ISAs and obfuscation levels on the new Bi...

  5. OASIF: An Efficient Obfuscation-Aware Self-Improving Framework for LLM-Based Assembly Code Instruction Following and Comprehension

    cs.SE 2026-06 unverdicted novelty 6.0

    OASIF improves open-source LLMs on obfuscated assembly comprehension by 5-17 percentage points on commercial VM obfuscators via a three-phase self-evolving training pipeline.

  6. Decaf: Improving Neural Decompilation with Automatic Feedback and Search

    cs.SE 2026-05 unverdicted novelty 6.0

    Decaf uses compiler feedback and search to improve neural decompilation, boosting semantic success rate from 26.0% to 83.9% on ExeBench Real -O2 split.

  7. CoDe-R: Refining Decompiler Output with LLMs via Rationale Guidance and Adaptive Inference

    cs.SE 2026-04 conditional novelty 6.0

    Rationale-guided fine-tuning plus dual-path adaptive inference lifts a 1.3B decompiler refiner to 50% average re-executability, a new lightweight SOTA on HumanEval-Decompile.

  8. CoDe-R: Refining Decompiler Output with LLMs via Rationale Guidance and Adaptive Inference

    cs.SE 2026-04 unverdicted novelty 6.0

    CoDe-R refines LLM decompiler output via rationale-guided semantic injection and dynamic fallback inference, making a 1.3B model the first to exceed 50% average re-executability on HumanEval-Decompile.

  9. Beyond the Edge of Function: Unraveling the Patterns of Type Recovery in Binary Code

    cs.CR 2025-03 unverdicted novelty 6.0

    ByteTR recovers variable types in binary code more effectively than prior methods by decoupling unbalanced type sets, mitigating compiler optimization effects via static analysis, and modeling inter-procedural data fl...

  10. When LLM Defenses Backfire: Characterizing Safety, Performance, and Cost Trade-offs

    cs.CR 2026-07 conditional novelty 5.0

    Across 11 defenses, 6 open LLMs, and multiple benchmarks, rule-based defenses best preserve task accuracy, conservative self-reflection drives over-refusal, and multi-round defenses dominate inference cost.

  11. LLM Agent-Assisted Reverse Engineering with Quantitative Readability Metrics

    cs.SE 2026-06 unverdicted novelty 5.0

    Presents the QRS composite metric (structural similarity gate plus Lexical Surprisal, Structural Simplicity, and Idiomatic Quality) to enable targeted LLM-agent improvements to decompiled code readability.

  12. Context-Guided Decompilation: A Step Towards Re-executability

    cs.SE 2025-11 unverdicted novelty 5.0

    ICL4Decomp applies in-context learning to guide LLMs in generating re-executable decompiled code from binaries, reporting roughly 40% higher re-executability than prior methods across datasets and optimization levels.