# Decaf: Improving Neural Decompilation with Automatic Feedback and Search

arXiv: https://arxiv.org/abs/2605.11501 · HTML: https://mlsanity.com/paper/2605.11501/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 54 references that are in the map.
- First posted on arXiv 12 May 2026.

## Builds on (most cited first, 24 of 24)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters](https://mlsanity.com/paper/2408.03314/) (2024)
- [Shortcut Learning in Deep Neural Networks](https://mlsanity.com/paper/2004.07780/) (2020)
- [Decision Transformer: Reinforcement Learning via Sequence Modeling](https://mlsanity.com/paper/2106.01345/) (2021)
- [Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation](https://mlsanity.com/paper/1609.08144/) (2016)
- [Competition-Level Code Generation with AlphaCode](https://mlsanity.com/paper/2203.07814/) (2022)
- [Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference](https://mlsanity.com/paper/1902.01007/) (2019)
- [Advantage-Weighted Regression: Simple and Scalable Off-Policy Reinforcement Learning](https://mlsanity.com/paper/1910.00177/) (2019)
- [Rethinking Benchmark and Contamination for Language Models with Rephrased Samples](https://mlsanity.com/paper/2311.04850/) (2023)
- [Controlling Linguistic Style Aspects in Neural Language Generation](https://mlsanity.com/paper/1707.02633/) (2017)
- [DIRE: A Neural Approach to Decompiled Identifier Naming](https://mlsanity.com/paper/1909.09029/) (2019)
- [SLaDe: A Portable Small Language Model Decompiler for Optimized Assembly](https://mlsanity.com/paper/2305.12520/) (2023)
- [EquiBench: Benchmarking Large Language Models' Reasoning about Program Semantics via Equivalence Checking](https://mlsanity.com/paper/2502.12466/) (2025)
- [Augmenting Decompiler Output with Learned Variable Names and Types](https://mlsanity.com/paper/2108.06363/) (2021)
- [Beyond the C: Retargetable Decompilation using Neural Machine Translation](https://mlsanity.com/paper/2212.08950/) (2022)
- [Polymorphic Type Inference for Machine Code](https://mlsanity.com/paper/1603.05495/) (2016)
- [BinSub: The Simple Essence of Polymorphic Type Inference for Machine Code](https://mlsanity.com/paper/2409.01841/) (2024)

## Built on by (most cited first, 1 of 1)

- [When LLM Decompilers Recompile More and Preserve Less](https://mlsanity.com/paper/2609.05370/) (2026)

## Abstract (arXiv)

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.