harite / mlsec

Research papers on ML for security

I collect several interesting research papers using machine learning for security.

Binary-related machine learning

1. [SEC'14] Byteweight: Learning to recognize functions in binary code (paper, project, slides, talk)
2. [SEC'15] Recognizing Functions in Binaries with Neural Networks (paper, talk)
3. [POPL'16] Estimating Types in Binaries using Predictive Modeling (paper, talk)
4. [SEC'17] Neural Nets Can Learn Function Type Signatures From Binaries (paper, project, talk)
5. [EUROSP'17] Compiler-Agnostic Function Detection in Binaries (paper, project, slides)
6. [CCS'17] Neural Network-based Graph Embedding for Cross-Platform Binary Code Similarity Detection (paper, slides)

Vulnerabilities detection

1. [CODASPY'16] Toward large-scale vulnerability discovery using Machine Learning (paper, slides, talk)
2. [ACSAC'17] Machine-Learning-Guided Typestate Analysis for Static Use-After-Free Detection (paper)

Symbolic execution

1. [NDSS'19] Neuro-Symbolic Execution: Augmenting Symbolic Execution with Neural Constraints (paper)

Taint analysis

1. [arxiv] Neutaint: Efficient Dynamic Taint Analysis with Neural Networks (paper)
2. [arxiv] Fine Grained Dataflow Tracking with Proximal Gradients (paper)

Testing deep learning system

1. [SOSP'17] DeepXplore: Automated Whitebox Testing of Deep Learning Systems (paper)
2. [ASE'18] DeepGauge: Multi-Granularity Testing Criteria for Deep Learning Systems (paper, project)
3. [ASE'18] Concolic Testing for Deep Neural Networks (paper, project)
4. [FSE'18] DLFuzz: Differential Fuzzing Testing of Deep Learning Systems (paper)
5. [PLMR'19] TensorFuzz: Debugging Neural Networks with Coverage-Guided Fuzzing (paper, project)
6. [ISSTA'19] DeepHunter: Hunting Deep Neural Network Defects via Coverage-Guided Fuzzing (paper)

Fuzzing

1. [ASE'17] Learn&Fuzz: Machine Learning for Input Fuzzing (paper)
2. [PAC'17] ExploitMeter: Combining Fuzzing with Machine Learning for Automated Evaluation of Software Exploitability (paper)
3. [arxiv] Not all bytes are equal: Neural byte sieve for fuzzing (paper)
4. [arxiv] Deep Reinforcement Fuzzing (paper)
5. [arxiv] FuzzerGym: A Competitive Framework for Fuzzing and Learning (paper)
6. [ISSTA'18] Compiler Fuzzing through Deep Learning (paper)
7. [SP'19] NEUZZ: Efficient Fuzzing with Neural Program Smoothing (paper, project, slides, talk)
8. [arxiv] A Review of Machine Learning Applications in Fuzzing (paper)
9. [SEC'19] EnFuzz: Ensemble Fuzzing with Seed Synchronization among Diverse Fuzzers (paper, project, talk)

Program analysis

1. [Keynote-CAV'19] Machine Learning as Program Synthesis (slides, blog: AI Safety as a PL problem)
2. [SEC'19] DEEPVSA: Facilitating Value-set Analysis with Deep Learning for Postmortem Program Analysis (paper, project, talk)