Training Language Model Agents to Find Vulnerabilities with CTF-Dojo
cs.SE, cs.CL, cs.CR, cs.LG
Submitted: 2025-08-25
Updated: 2026-11-13
Code: https://github.com/amazon-science/CTF-Dojo
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Purple Llama CyberSecEval: A Secure Coding Benchmark for Language Models
- AutoAdvExBench: Benchmarking autonomous exploitation of adversarial example defenses
- From Text to MITRE Techniques: Exploring the Malicious Use of Large Language Models for Generating Cyber Attack Payloads
- Evaluating Large Language Models Trained on Code
- Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
- SecRepoBench: Benchmarking Code Agents for Secure Code Completion in Real-World Repositories
- LLM Agents can Autonomously Hack Websites
- The Llama 3 Herd of Models
- Qwen2.5-Coder Technical Report
- GPT-4o System Card
- OpenAI o1 System Card
- NeMo: a toolkit for building AI applications using Neural Modules
- SEC-bench: Automated Benchmarking of LLM Agents on Real-World Software Security Tasks
- StarCoder: may the source be with you!
- DeepSeek-V3 Technical Report
- SecQA: A Concise Question-Answering Dataset for Evaluating Large Language Models in Computer Security
- StarCoder 2 and The Stack v2: The Next Generation
- An Adversarial Perspective on Machine Unlearning for AI Safety
- Lingma SWE-GPT: An Open Development-Process-Centric Language Model for Automated Software Improvement
- CAI: An Open, Bug Bounty-Ready Cybersecurity AI
Related papers
- ArkEval: Benchmarking and Evaluating Automated CodeRepair for ArkTS
- TestPrism: Rethinking Test Evaluation Beyond a Single Reference
- Mining Workflow Graphs for Black-Box Boundary Testing of Conversational LLM Agents
- Falsification-Based Verification of LLM-Generated Optimization Models: Sound Test Batteries and Their Detection Limits
- GitSkills: A Dataset of Agent Skills on GitHub
- For Coding Agents, the Repository Matters