Summary of Belief-enriched Pessimistic Q-learning Against Adversarial State Perturbations, by Xiaolin Sun et al.
Belief-Enriched Pessimistic Q-Learning against Adversarial State Perturbations
by Xiaolin Sun, Zizhan Zheng
First submitted to arxiv on: 6 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary In this paper, researchers explore the vulnerability of reinforcement learning (RL) agents to attacks that manipulate their state observations. Current solutions either regularize the trained policy or train both the agent’s and attacker’s policies, but these approaches have limitations. The proposed algorithm derives a pessimistic policy to mitigate uncertainty about true states, combined with belief state inference and diffusion-based state purification. This approach demonstrates excellent performance under strong attacks while maintaining a comparable training overhead. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Reinforcement learning has achieved great success in many areas, but it’s also very vulnerable to attacks that change what the agent sees. Some people have tried to fix this by adding a special term to the agent’s training or by training both the good guy and bad guy’s policies. But these ideas aren’t perfect. This paper introduces a new way for reinforcement learning agents to be more careful about what they see, which helps them work well even when someone is trying to trick them. |
Keywords
* Artificial intelligence * Diffusion * Inference * Reinforcement learning
