mlsanityData to 7 Oct 2026

QiMeng-NeuComBack: Self-Evolving Translation from IR to Assembly Code

arXiv 2511.01183 ↗First posted 3 Nov 2025cs.AI cs.PLcited by 0

On the map: Synthetic data and self-improvement for LLMs → Code and instruction data synthesis

Abstract (arXiv)

Compilers, while essential, are notoriously complex systems that demand prohibitively expensive human expertise to develop and maintain. The recent advancements in Large Language Models (LLMs) offer a compelling new paradigm: Neural Compilation, which could potentially simplify compiler development for new architectures and facilitate the discovery of innovative optimization techniques. However, several critical obstacles impede its practical adoption. Firstly, a significant lack of dedicated benchmarks and robust evaluation methodologies hinders objective assessment and tracking of progress in the field. Secondly, systematically enhancing the reliability and performance of LLM-generated assembly remains a critical challenge. Addressing these challenges, this paper introduces NeuComBack, a novel benchmark dataset specifically designed for IR-to-assembly compilation. Leveraging this dataset, we first define a foundational Neural Compilation workflow and conduct a comprehensive evaluation of the capabilities of recent frontier LLMs on Neural Compilation, establishing new performance baselines. We further propose a self-evolving prompt optimization method that enables LLMs to iteratively evolve their internal prompt strategies by extracting insights from prior self-debugging traces, thereby enhancing their neural compilation capabilities. Experiments demonstrate that our method significantly improves both the functional correctness and the performance of LLM-generated assembly code. Compared to baseline prompts, the functional correctness rates improved from 44% to 64% on x86_64 and from 36% to 58% on aarch64, respectively. More significantly, among the 16 correctly generated x86_64 programs using our method, 14 (87.5%) surpassed clang-O3 performance.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Code and instruction data synthesis, one of 13 directions in Synthetic data and self-improvement for LLMs. Builds on 32 of its 52 references that are in the map. First posted on arXiv 3 Nov 2025.

Builds on

32 of 52 references.

References · 32

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2024The Llama 3 Herd of Modelscited by 14,192
  3. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  4. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,601
  5. 2024Qwen2.5 Technical Reportcited by 4,713
  6. 2024GPT-4o System Cardcited by 4,467
  7. 2022Large Language Models Are Human-Level Prompt Engineerscited by 1,020
  8. 2022Few-Shot Parameter-Efficient Fine-Tuning is Better and Cheaper than In-Context Learningcited by 886
  9. 2023Large Language Models as Optimizerscited by 830
  10. 2022Automatic Chain of Thought Prompting in Large Language Modelscited by 684
  11. 2023Automatic Prompt Optimization with "Gradient Descent" and Beam Searchcited by 628
  12. 2023Large Language Models for Software Engineering: A Systematic Literature Reviewcited by 389
20 more
  1. 2023ChatIE: Zero-Shot Information Extraction via Chatting with ChatGPTcited by 250
  2. 2023Software Testing with Large Language Models: Survey, Landscape, and Visioncited by 163
  3. 2023Automatic Prompt Augmentation and Selection with Chain-of-Thought from Labeled Datacited by 119
  4. 2022Large Language Models are Zero-Shot Fuzzers: Fuzzing Deep-Learning Libraries via Large Language Modelscited by 102
  5. 2023Can ChatGPT Defend its Belief in Truth? Evaluating LLM Reasoning via Debatecited by 92
  6. 2023Can ChatGPT replace StackOverflow? A Study on Robustness and Reliability of Large Language Model Code Generationcited by 72
  7. 2023Large Language Models for Compiler Optimizationcited by 52
  8. 2021MLGO: a Machine Learning Guided Compiler Optimizations Frameworkcited by 40
  9. 2024Meta Large Language Model Compiler: Foundation Models of Compiler Optimizationcited by 39
  10. 2024LLM4Decompile: Decompiling Binary Code with Large Language Modelscited by 38
  11. 2023WhiteFox: White-Box Compiler Fuzzing Empowered by Large Language Modelscited by 37
  12. 2023Enhancing Zero-Shot Chain-of-Thought Reasoning in Large Language Models through Logiccited by 34
  13. 2023AutoHint: Automatic Prompt Optimization with Hint Generationcited by 29
  14. 2023SLaDe: A Portable Small Language Model Decompiler for Optimized Assemblycited by 28
  15. 2023Refining Decompiled C Code with Large Language Modelscited by 24
  16. 2025The Prompt Alchemist: Automated LLM-Tailored Prompt Optimization for Test Case Generationcited by 16
  17. 2022Beyond the C: Retargetable Decompilation using Neural Machine Translationcited by 15
  18. 2018Machine Learning in Compiler Optimisationcited by 13
  19. 2021Learning C to x86 Translation: An Experiment in Neural Compilationcited by 12
  20. 2025Prompt Alchemy: Automatic Prompt Refinement for Enhancing Code Generationcited by 8

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024Should AI Optimize Your Code? A Comparative Study of Classical Optimizing Compilers Versus Current Large Language Models
  2. 2025REASONING COMPILER: LLM-Guided Optimizations for Efficient Model Serving
  3. 2026From Brute Force to Semantic Insight: Performance-Guided Data Transformation Design with LLMs
  4. 2026An MLIR-Based Compilation Method for Large Language Models
  5. 2025Can Large Language Models Understand Intermediate Representations in Compilers?
  6. 2023LLM4VV: Developing LLM-Driven Testsuite for Compiler Validation
  7. 2024CompilerDream: Learning a Compiler World Model for General Code Optimization
  8. 2026AutoPass: Evidence-Guided LLM Agents for Compiler Performance Tuning