Research map / Adversarial robustness and security / Attacks on text and graphs
Actor Critic Deep Reinforcement Learning for Neural Malware Control
Yu Wang, Jack Stokes, Mady Marinescu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace DQN with actor critic with adversarial-modified critic | emulation halting model · early halting, N=2000, K=20 | % files halted early: 91.3% -> 93.6% | Improved |
| replace DQN with actor critic | emulation halting model · malware detection at FPR 1% | true positive rate: 69.5% -> 76.4% | Improved |
| replace LSTM file classifier with actor critic halting model | malware classifier · malware detection at FPR 1% | true positive rate: 41.2% -> 76.4% | Improved |
| increase number of training files N from 30 to 2000 (ablation) | actor critic halting · early halting, K=20 | % files halted early: 63.5% -> 93.6% | Improved |
| increase required consecutive halt actions K from 10 to 20 (ablation) | actor critic halting · early halting, N=2000 | % files halted early: 99.4% -> 93.6% | Hurt |
Closest papers by meaning
- Adversarial Inception Backdoor Attacks against Reinforcement Learning (ICML 2025 · similarity 0.81)
- PolicyGuard: Towards Test-time and Step-level Adversary Defense for Reinforcement Learning Agent (ICML 2026 · similarity 0.81)
- Breaking the Barrier: Enhanced Utility and Robustness in Smoothed DRL Agents (ICML 2024 · similarity 0.81)
- Confounding Robust Deep Reinforcement Learning: A Causal Approach (NeurIPS 2025 · similarity 0.81)
- Exploring the Efficacy of Multi-Agent Reinforcement Learning for Autonomous Cyber Defence: A CAGE Challenge 4 Perspective (AAAI 2025 · similarity 0.81)
- Reinforcement Learning Platform for Adversarial Black-box Attacks with Custom Distortion Filters (AAAI 2025 · similarity 0.81)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
