PoC-Adapt: Semantic-Aware Automated Vulnerability Reproduction with LLM Multi-Agents and Reinforcement Learning-Driven Adaptive Policy

• URL: http://arxiv.org/abs/2604.06618v2
• Date: Thu, 09 Apr 2026 14:37:23 GMT
• Title: PoC-Adapt: Semantic-Aware Automated Vulnerability Reproduction with LLM Multi-Agents and Reinforcement Learning-Driven Adaptive Policy
• Abstract summary: We present PoC-Adapt, an end-to-end framework for automated PoC generation and verification.<n>PoC-Adapt is architected upon a foundation semantic runtime validation and adaptive policy learning.<n>Experiments show that PoC-Adapt significantly improves verification reliability by 25% and reduces exploit generation cost.
• Score: 0.47998222538650537
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: While recent approaches leverage large language models (LLMs) and multi-agent pipelines to automatically generate proof-of-concept (PoC) exploits from vulnerability reports, existing systems often suffer from two fundamental limitations: unreliable validation based on surface-level execution signals and high operational cost caused by extensive trial-and-error during exploit generation. In this paper, we present PoC-Adapt, an end-to-end framework for automated PoC generation and verification, architected upon a foundation semantic runtime validation and adaptive policy learning. At the core of PoC-Adapt is a Semantic Oracle that validates exploits by comparing structured pre- and post-execution system states, enabling reliable distinction between true vulnerability exploitation and incidental behavioral changes. To reduce exploration cost, we further introduce an Adaptive Policy Learning mechanism that learns an exploitation policy over semantic states and actions, guiding the exploit agent toward effective strategies with fewer failed attempts. PoC-Adapt is implemented as a multi-agent system comprising specialized agents for root cause analysis, environment building, exploit generation, and semantic validation, coordinated through structured feedback loops. Experimenting on the CWE-Bench-Java and PrimeVul benchmarks shows that PoC-Adapt significantly improves verification reliability by 25% and reduces exploit generation cost compared to prior LLM-based systems, highlighting the importance of semantic validation and learned action policies in automated vulnerability reproduction. Applied to the latest CVE corpus, PoC-Adapt confirmed 12 verified PoC out of 80 reproduce attempts at a cost of $0.42 per generated exploit

Related papers

• OGPO: Sample Efficient Full-Finetuning of Generative Control Policies [53.42266064673132]
  Generative control policies (GCPs) have emerged as effective parameterizations for robot learning.<n>This work introduces Off-policy Generative Policy Optimization (OGPO), a sample-efficient for finetuning GCPs.<n>OGPO achieves state-of-the-art performance on manipulation tasks spanning multi-task settings, high-precision insertion, and dexterous control.
  arXiv  Detail & Related papers  (2026-05-04T18:36:40Z)
• Program Analysis Guided LLM Agent for Proof-of-Concept Generation [3.462916206368371]
  We present a novel approach called Program Analysis Guided proof of concept generation agENT (PAGENT)<n>PAGENT integrates lightweight and rule-based static analysis phases for providing static analysis guidance and sanitizer-based profiling and coverage information for providing dynamic analysis guidance with a PoC generation agent.<n>Our experiments demonstrate that the resulting hybrid approach significantly outperforms the prior top-performing agentic approach by 132% for the PoC generation task.
  arXiv  Detail & Related papers  (2026-04-08T21:55:18Z)
• SecPI: Secure Code Generation with Reasoning Models via Security Reasoning Internalization [50.71047638695205]
  Reasoning language models (RLMs) are increasingly used in programming.<n>Yet, even state-of-the-art RLMs frequently introduce critical security vulnerabilities in generated code.<n>We present SecPI, a fine-tuning pipeline that teaches RLMs to internalize structured security reasoning.
  arXiv  Detail & Related papers  (2026-04-04T04:29:11Z)
• CaP-X: A Framework for Benchmarking and Improving Coding Agents for Robot Manipulation [48.85772216740915]
  "Code-as-Policy" considers how executable code can complement data-intensive Vision-Language-Action methods.<n>We present CaP-X, an open-access framework for studying Code-as-Policy agents in robot manipulation.
  arXiv  Detail & Related papers  (2026-03-23T18:08:10Z)
• Execution-State-Aware LLM Reasoning for Automated Proof-of-Vulnerability Generation [36.950993500170014]
  We present DrillAgent, an agentic framework that reformulates PoV generation as an iterative hypothesis-verification-refinement process.<n>We evaluate DrillAgent on SEC-bench, a large-scale benchmark of real-world C/C++ vulnerabilities.
  arXiv  Detail & Related papers  (2026-02-14T03:17:27Z)
• Co-RedTeam: Orchestrated Security Discovery and Exploitation with LLM Agents [57.49020237126194]
  Large language models (LLMs) have shown promise in assisting cybersecurity tasks, yet existing approaches struggle with automatic vulnerability discovery and exploitation.<n>We propose Co-RedTeam, a security-aware multi-agent framework designed to mirror real-world red-teaming.<n>Co-RedTeam decomposes vulnerability analysis into coordinated discovery and exploitation stages, enabling agents to plan, execute, validate, and refine actions.
  arXiv  Detail & Related papers  (2026-02-02T14:38:45Z)
• An Empirical Evaluation of LLM-Based Approaches for Code Vulnerability Detection: RAG, SFT, and Dual-Agent Systems [1.5216960763930782]
  The rapid advancement of Large Language Models (LLMs) presents new opportunities for automated software vulnerability detection.<n>This paper presents a comparative study on the effectiveness of LLM-based techniques for detecting software vulnerabilities.
  arXiv  Detail & Related papers  (2026-01-01T08:05:51Z)
• Metacognitive Self-Correction for Multi-Agent System via Prototype-Guided Next-Execution Reconstruction [58.51530390018909]
  Large Language Model based multi-agent systems excel at collaborative problem solving but remain brittle to cascading errors.<n>We present MASC, a metacognitive framework that endows MAS with real-time, unsupervised, step-level error detection and self-correction.
  arXiv  Detail & Related papers  (2025-10-16T05:35:37Z)
• VulAgent: Hypothesis-Validation based Multi-Agent Vulnerability Detection [55.957275374847484]
  VulAgent is a multi-agent vulnerability detection framework based on hypothesis validation.<n>It implements a semantics-sensitive, multi-view detection pipeline, each aligned to a specific analysis perspective.<n>On average, VulAgent improves overall accuracy by 6.6%, increases the correct identification rate of vulnerable--fixed code pairs by up to 450%, and reduces the false positive rate by about 36%.
  arXiv  Detail & Related papers  (2025-09-15T02:25:38Z)
• VulnRepairEval: An Exploit-Based Evaluation Framework for Assessing Large Language Model Vulnerability Repair Capabilities [41.85494398578654]
  VulnRepairEval is an evaluation framework anchored in functional Proof-of-Concept exploits.<n>Our framework delivers a comprehensive, containerized evaluation pipeline that enables reproducible differential assessment.
  arXiv  Detail & Related papers  (2025-09-03T14:06:10Z)
• A Generative Framework for Low-Cost Result Validation of Machine Learning-as-a-Service Inference [4.478182379059458]
  Fides is a novel framework for real-time integrity validation of ML-as-a-Service (ML) inference. Fides features a client-side attack detection model that uses statistical analysis and divergence measurements to identify, with a high likelihood, if the service model is under attack. We devised a generative adversarial network framework for training the attack detection and re-classification models.
  arXiv  Detail & Related papers  (2023-03-31T19:17:30Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.