# Top Score on the Wrong Exam: On Benchmarking in Machine Learning for Vulnerability Detection

arXiv: https://arxiv.org/abs/2408.12986 · HTML: https://mlsanity.com/paper/2408.12986/

## Key facts

- Cited by 28 papers in the map (data through 14 Oct 2026).
- Builds on 28 of its 66 references that are in the map.
- First posted on arXiv 23 Aug 2024.

## Builds on (most cited first, 28 of 28)

- [CodeT5: Identifier-aware Unified Pre-trained Encoder-Decoder Models for Code Understanding and Generation](https://mlsanity.com/paper/2109.00859/) (2021)
- [CodeXGLUE: A Machine Learning Benchmark Dataset for Code Understanding and Generation](https://mlsanity.com/paper/2102.04664/) (2021)
- [UniXcoder: Unified Cross-Modal Pre-training for Code Representation](https://mlsanity.com/paper/2203.03850/) (2022)
- [An Investigation of Why Overparameterization Exacerbates Spurious Correlations](https://mlsanity.com/paper/2005.04345/) (2020)
- [Devign: Effective Vulnerability Identification by Learning Comprehensive Program Semantics via Graph Neural Networks](https://mlsanity.com/paper/1909.03496/) (2019)
- [On Feature Learning in the Presence of Spurious Correlations](https://mlsanity.com/paper/2210.11369/) (2022)
- [Deep Learning based Vulnerability Detection: Are We There Yet?](https://mlsanity.com/paper/2009.07235/) (2020)
- [DiverseVul: A New Vulnerable Source Code Dataset for Deep Learning Based Vulnerability Detection](https://mlsanity.com/paper/2304.00409/) (2023)
- [LLMs Cannot Reliably Identify and Reason About Security Vulnerabilities (Yet?): A Comprehensive Evaluation, Framework, and Benchmarks](https://mlsanity.com/paper/2312.12575/) (2023)
- [Vulnerability Detection with Fine-grained Interpretations](https://mlsanity.com/paper/2106.10478/) (2021)
- [Data Quality for Software Vulnerability Datasets](https://mlsanity.com/paper/2301.05456/) (2023)
- [An Empirical Study of Deep Learning Models for Vulnerability Detection](https://mlsanity.com/paper/2212.08109/) (2022)
- [A Deep Dive into Large Language Models for Automated Bug Localization and Repair](https://mlsanity.com/paper/2404.11595/) (2024)
- [Uncovering the Limits of Machine Learning for Automatic Vulnerability Detection](https://mlsanity.com/paper/2306.17193/) (2023)
- [Dataflow Analysis-Inspired Deep Learning for Efficient Vulnerability Detection](https://mlsanity.com/paper/2212.08108/) (2022)
- [Pre-training by Predicting Program Dependencies for Vulnerability Analysis Tasks](https://mlsanity.com/paper/2402.00657/) (2024)
- [SCALE: Constructing Structured Natural Language Comment Trees for Software Vulnerability Detection](https://mlsanity.com/paper/2403.19096/) (2024)
- [Graph Neural Networks for Vulnerability Detection: A Counterfactual Explanation](https://mlsanity.com/paper/2404.15687/) (2024)
- [Distinguishing Look-Alike Innocent and Vulnerable Code by Subtle Semantic Representation Learning and Explanation](https://mlsanity.com/paper/2308.11237/) (2023)
- [On the Effectiveness of Function-Level Vulnerability Detectors for Inter-Procedural Vulnerabilities](https://mlsanity.com/paper/2401.09767/) (2024)
- [Towards Causal Deep Learning for Vulnerability Detection](https://mlsanity.com/paper/2310.07958/) (2023)
- [Toward Improved Deep Learning-based Vulnerability Detection](https://mlsanity.com/paper/2403.03024/) (2024)
- [Vulnerability Detection with Graph Simplification and Enhanced Graph Representation Learning](https://mlsanity.com/paper/2302.04675/) (2023)
- [When Less is Enough: Positive and Unlabeled Learning Model for Vulnerability Detection](https://mlsanity.com/paper/2308.10523/) (2023)
- [Coca: Improving and Explaining Graph Neural Network-Based Vulnerability Detection Systems](https://mlsanity.com/paper/2401.14886/) (2024)
- [VulSPG: Vulnerability detection based on slice property graph representation learning](https://mlsanity.com/paper/2109.02527/) (2021)
- [CONCORD: Clone-aware Contrastive Learning for Source Code](https://mlsanity.com/paper/2306.03234/) (2023)
- [Semantic Fuzzing with Zest](https://mlsanity.com/paper/1812.00078/) (2018)

## Built on by (most cited first, 28 of 28)

- [RepoAudit: An Autonomous LLM-Agent for Repository-Level Code Auditing](https://mlsanity.com/paper/2501.18160/) (2025)
- [Benchmarking LLMs and LLM-based Agents in Practical Vulnerability Detection for Code Repositories](https://mlsanity.com/paper/2503.03586/) (2025)
- [From CVE Entries to Verifiable Exploits: An Automated Multi-Agent Framework for Reproducing CVEs](https://mlsanity.com/paper/2509.01835/) (2025)
- [SecureVibeBench: Benchmarking Secure Vibe Coding of AI Agents via Reconstructing Vulnerability-Introducing Scenarios](https://mlsanity.com/paper/2509.22097/) (2025)
- [VulnLLM-R: Specialized Reasoning LLM with Agent Scaffold for Vulnerability Detection](https://mlsanity.com/paper/2512.07533/) (2025)
- [Locus: Agentic Predicate Synthesis for Directed Fuzzing](https://mlsanity.com/paper/2508.21302/) (2025)
- [A Systematic Literature Review on Detecting Software Vulnerabilities with Large Language Models](https://mlsanity.com/paper/2507.22659/) (2025)
- [Assessing the Impact of Code Changes on the Fault Localizability of Large Language Models](https://mlsanity.com/paper/2504.04372/) (2025)
- [LLM-based Vulnerability Discovery through the Lens of Code Metrics](https://mlsanity.com/paper/2509.19117/) (2025)
- [From SFT to RL: Demystifying the Post-Training Pipeline for LLM-based Vulnerability Detection](https://mlsanity.com/paper/2602.14012/) (2026)
- [DREA: Decoupled Reasoning and Exploration Agents for Repository-Level Vulnerability Detection](https://mlsanity.com/paper/2607.13439/) (2026)
- [BlockScan: Detecting Anomalies in Blockchain Transactions](https://mlsanity.com/paper/2410.04039/) (2024)
- [An Empirical Study of Vulnerabilities in Python Packages and Their Detection](https://mlsanity.com/paper/2509.04260/) (2025)
- [Repairing vulnerabilities without invisible hands. A differentiated replication study on LLMs](https://mlsanity.com/paper/2507.20977/) (2025)
- [Toward Scalable Automated Repository-Level Datasets for Software Vulnerability Detection](https://mlsanity.com/paper/2603.17974/) (2026)
- [LLM-Based Agents for Identifying Bug-Introducing Commits](https://mlsanity.com/paper/2603.29378/) (2026)
- [Direction for Detection: A Survey of Automated Vulnerability Detection and all of its Pain Points](https://mlsanity.com/paper/2412.11194/) (2024)
- [On the Difficulty of Selecting Few-Shot Examples for Effective LLM-based Vulnerability Detection](https://mlsanity.com/paper/2510.27675/) (2025)
- [VulStyle: A Multi-Modal Pre-Training for Code Stylometry-Augmented Vulnerability Detection](https://mlsanity.com/paper/2604.26313/) (2026)
- [A Comprehensive Evaluation of Code Language Models for Security Patch Detection](https://mlsanity.com/paper/2605.13138/) (2026)
- [Evaluating LLMs for Real-World Web Vulnerability Detection](https://mlsanity.com/paper/2606.21397/) (2026)
- [ExplainBench: Evaluating Code Explanations from Agents](https://mlsanity.com/paper/2607.26451/) (2026)
- [CyberForge: Verified Vulnerability Injection at Repository Level for Cybersecurity Agent Training](https://mlsanity.com/paper/2608.06471/) (2026)
- [On Predicting Vulnerability Severity Using In-Context Learning: An Industrial Case Study](https://mlsanity.com/paper/2608.22089/) (2026)
- [Plug 'n' Pray: Agentic LLM-based Detection of Potential Log File Exposures in Third-Party Content Management System Plugins](https://mlsanity.com/paper/2609.17164/) (2026)
- [SelfOp: An Optimization Algorithm for Self-Improving Security Agents](https://mlsanity.com/paper/2609.22792/) (2026)
- [A Function-Level Vulnerability Score Measures Flag Rate More Than the Model: Protocol Effects on Paired Benchmarks](https://mlsanity.com/paper/2609.32890/) (2026)
- [VulValidate: Auditing Function-Level Vulnerability Labels with Executable Evidence](https://mlsanity.com/paper/2610.05103/) (2026)

## Abstract (arXiv)

According to our survey of machine learning for vulnerability detection (ML4VD), 9 in every 10 papers published in the past five years define ML4VD as a function-level binary classification problem: Given a function, does it contain a security flaw? From our experience as security researchers, faced with deciding whether a given function makes the program vulnerable to attacks, we would often first want to understand the context in which this function is called. In this paper, we study how often this decision can really be made without further context and study both vulnerable and non-vulnerable functions in the most popular ML4VD datasets. We call a function "vulnerable" if it was involved in a patch of an actual security flaw and confirmed to cause the program's vulnerability. It is "non-vulnerable" otherwise. We find that in almost all cases this decision cannot be made without further context. Vulnerable functions are often vulnerable only because a corresponding vulnerability-inducing calling context exists while non-vulnerable functions would often be vulnerable if a corresponding context existed. But why do ML4VD techniques achieve high scores even though there is demonstrably not enough information in these samples? Spurious correlations: We find that high scores can be achieved even when only word counts are available. This shows that these datasets can be exploited to achieve high scores without actually detecting any security vulnerabilities. We conclude that the prevailing problem statement of ML4VD is ill-defined and call into question the internal validity of this growing body of work. Constructively, we call for more effective benchmarking methodologies to evaluate the true capabilities of ML4VD, propose alternative problem statements, and examine broader implications for the evaluation of machine learning and programming analysis research.