Research map / Adversarial robustness and security / Attacks on text and graphs
Stealthy and Efficient Adversarial Attacks against Deep Reinforcement Learning
Jianwen Sun, Tianwei Zhang, Xiaofei Xie, Lei Ma, Yan Zheng, Kangjie Chen, Yang Liu
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 |
|---|---|---|---|
| Use Critical Point Attack to select attack moments using predicted long-term damage | A3C · Pong | attack steps: 6 steps for Pong and 2 steps | Improved |
| Use Critical Point Attack to select attack moments using predicted long-term damage | A3C · Breakout | attack steps: 5 steps for Breakout and 2 steps | Improved |
| Use Critical Point Attack to select attack moments using predicted long-term damage | DDPG · TORCS | attack steps: 80 steps; 1 step | Improved |
| Use Antagonist Attack to learn attack timing and target actions | A3C · Pong | attack steps: 6 steps for Pong and 3 steps | Improved |
| Use Antagonist Attack to learn attack timing and target actions | A3C · Breakout | attack steps: 5 steps for Breakout and 3 steps | Improved |
| Use Antagonist Attack to learn attack timing and target actions | PPO · InvertedPendulum | attack steps: 2 | Improved |
| Use Antagonist Attack to learn attack timing and target actions | PPO · Hopper | attack steps: 1 | Improved |
| Use Antagonist Attack to learn attack timing and target actions | PPO · HalfCheetah | attack steps: 5 | Improved |
| Use Antagonist Attack to learn attack timing and target actions | PPO · Walker2d | attack steps: 3 | Improved |
| Use Critical Point Attack instead of Antagonist Attack | A3C · Pong and Breakout | attack steps: Antagonist Attack ... 3 steps; Critical Point Attack (2 steps) | Improved |
Closest papers by meaning
- Adversarial Inception Backdoor Attacks against Reinforcement Learning (ICML 2025 · similarity 0.90)
- Rethinking Adversarial Policies: A Generalized Attack Formulation and Provable Defense in RL (ICLR 2024 · similarity 0.89)
- Diffusion Guided Adversarial State Perturbations in Reinforcement Learning (NeurIPS 2025 · similarity 0.88)
- RAT: Adversarial Attacks on Deep Reinforcement Agents for Targeted Behaviors (AAAI 2025 · similarity 0.87)
- Beyond Worst-case Attacks: Robust RL with Adaptive Defense via Non-dominated Policies (ICLR 2024 · similarity 0.87)
- Optimal Attack and Defense for Reinforcement Learning (AAAI 2024 · similarity 0.87)
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.
