Lune

ISSTA2023Top-tier venue

Large Language Models Are Zero-Shot Fuzzers: Fuzzing Deep-Learning Libraries via Large Language Models

Yinlin Deng, Chunqiu Steven Xia, Haoran Peng, Chenyuan Yang, Lingming Zhang

Abstract

Deep Learning (DL) systems have received exponential growth in popularity and have become ubiquitous in our everyday life. Such systems are built on top of popular DL libraries, e.g., TensorFlow and PyTorch which provide APIs as building blocks for DL systems. Detecting bugs in these DL libraries is critical for almost all downstream DL systems in ensuring effectiveness/safety for end users. Meanwhile, traditional fuzzing techniques can be hardly effective for such a challenging domain since the input DL programs need to satisfy both the input language (e.g., Python) syntax/semantics and the DL API input/shape constraints for tensor computations. To address these limitations, we propose TitanFuzz – the first approach to directly leveraging Large Language Models (LLMs) to generate input programs for fuzzing DL libraries. LLMs are titanic models trained on billions of code snippets and can autoregressively generate human-like code snippets. Our key insight is that modern LLMs can also include numerous code snippets invoking DL library APIs in their training corpora, and thus can implicitly learn both language syntax/semantics and intricate DL API constraints for valid DL program generation. More specifically, we use both generative and infilling LLMs (e.g., Codex/InCoder) to generate and mutate valid/diverse input DL programs for fuzzing. Our experimental results demonstrate that TitanFuzz can achieve 30.38%/50.84% higher code coverage than state-of-the-art fuzzers on TensorFlow/PyTorch. Furthermore, TitanFuzz is able to detect 65 bugs, with 44 already confirmed as previously unknown bugs. This paper demonstrates that modern titanic LLMs can be leveraged to directly perform both generation-based and mutation-based fuzzing studied for decades, while being fully automated, generalizable, and applicable to domains challenging for traditional approaches (such as DL systems). We hope TitanFuzz can stimulate more work in this promising direction of LLMs for fuzzing.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

101 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • SWE-RL: Advancing LLM Reasoning via Reinforcement Learning on Open Software Evolution cites Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generation
  • SWE-RL: Advancing LLM Reasoning via Reinforcement Learning on Open Software Evolution cites Magicoder: Empowering Code Generation with OSS-Instruct
  • SWE-RL: Advancing LLM Reasoning via Reinforcement Learning on Open Software Evolution cites Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
  • Fuzz4All: Universal Fuzzing with Large Language Models cites CodeBERT: A Pre-Trained Model for Programming and Natural Languages
  • Fuzz4All: Universal Fuzzing with Large Language Models cites Language Models are Few-Shot Learners
  • Fuzz4All: Universal Fuzzing with Large Language Models cites The Curious Case of Neural Text Degeneration
  • Language Models are Few-Shot Learners cites The Curious Case of Neural Text Degeneration
  • Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generation cites Language Models are Few-Shot Learners
  • GPTScan: Detecting Logic Vulnerabilities in Smart Contracts by Combining GPT with Program Analysis cites Language Models are Few-Shot Learners
  • Magicoder: Empowering Code Generation with OSS-Instruct cites Language Models are Few-Shot Learners
  • Magicoder: Empowering Code Generation with OSS-Instruct cites Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generation
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers101

  1. Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code GenerationJiawei Liu, Chunqiu Steven Xia, Yuyao Wang, Lingming ZhangNeurIPS 2023 · 2,317 citations
  2. SWE-RL: Advancing LLM Reasoning via Reinforcement Learning on Open Software EvolutionYuxiang Wei, Olivier Duchenne, Jade Copet, Quentin Carbonneaux et al.NeurIPS 2025 · 291 citations
  3. Magicoder: Empowering Code Generation with OSS-InstructYuxiang Wei, Zhe Wang, Jiawei Liu, Yifeng Ding et al.ICML 2024 · 246 citations
  4. Fuzz4All: Universal Fuzzing with Large Language ModelsChunqiu Steven Xia, Matteo Paltenghi, Jia Le Tian, Michael Pradel et al.ICSE 2024 · 155 citations
  5. GPTScan: Detecting Logic Vulnerabilities in Smart Contracts by Combining GPT with Program AnalysisYuqiang Sun, Daoyuan Wu, Yue Xue, Han Liu et al.ICSE 2024 · 131 citations

Builds on28

  1. Language Models are Few-Shot LearnersTom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah et al.NeurIPS 2020 · 64,255 citations
  2. Chain-of-Thought Prompting Elicits Reasoning in Large Language ModelsJason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma et al.NeurIPS 2022 · 22,562 citations
  3. The Curious Case of Neural Text DegenerationAri Holtzman, Jan Buys, Li Du, Maxwell Forbes et al.ICLR 2020 · 4,112 citations
  4. CodeBERT: A Pre-Trained Model for Programming and Natural LanguagesZhangyin Feng, Daya Guo, Duyu Tang, Nan Duan et al.EMNLP Findings 2020 · 2,828 citations
  5. Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code GenerationJiawei Liu, Chunqiu Steven Xia, Yuyao Wang, Lingming ZhangNeurIPS 2023 · 2,317 citations

Related papers

  1. Large Language Models are Edge-Case Generators: Crafting Unusual Programs for Fuzzing Deep Learning LibrariesYinlin Deng, Chunqiu Steven Xia, Chenyuan Yang, Shizhuo Dylan Zhang et al.ICSE 2024 · 85 citations
  2. Your Fix Is My Exploit: Enabling Comprehensive DL Library API Fuzzing with Large Language ModelsKunpeng Zhang, Shuai Wang, Jitao Han, Xiaogang Zhu et al.ICSE 2025 · 6 citations
  3. Free Lunch for Testing: Fuzzing Deep-Learning Libraries from Open SourceAnjiang Wei, Yinlin Deng, Chenyuan Yang, Lingming ZhangICSE 2022 · 91 citations
  4. The Seeds of the Future Sprout from History: Fuzzing for Unveiling Vulnerabilities in Prospective Deep-Learning LibrariesZhiyuan Li, Jingzheng Wu, Xiang Ling, Tianyue Luo et al.ICSE 2025 · 4 citations
  5. NeuRI: Diversifying DNN Generation via Inductive Rule InferenceJiawei Liu, Jinjun Peng, Yuyao Wang, Lingming ZhangFSE 2023 · 24 citations
Dusk over the sea between two cliffs drawn in fine vertical lines