Summary of Towards Robust Policy: Enhancing Offline Reinforcement Learning with Adversarial Attacks and Defenses, by Thanh Nguyen et al.
Towards Robust Policy: Enhancing Offline Reinforcement Learning with Adversarial Attacks and Defenses
by Thanh Nguyen, Tung M. Luu, Tri Ton, Chang D. Yoo
First submitted to arxiv on: 18 May 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Robotics (cs.RO)
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 The proposed framework enhances the robustness of offline reinforcement learning (RL) models by leveraging advanced adversarial attacks and defenses. Offline RL addresses the challenge of expensive data exploration, but this training paradigm can compromise policy robustness due to observation perturbations or intentional attacks. The framework attacks the actor and critic components by perturbing observations during training and uses adversarial defenses as regularization to enhance the learned policy. Four attacks and two defenses are introduced and evaluated on the D4RL benchmark, showing the vulnerability of both the actor and critic to attacks and the effectiveness of the defenses in improving policy robustness. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Offline RL helps train policies using vast amounts of data, but this can make them vulnerable to real-world conditions. A new framework makes offline RL models more reliable by fighting back against fake observations during training. The approach uses four different types of attacks on the model and two ways to defend against these attacks. This is tested on a dataset called D4RL, which shows that the attacks work well and the defenses can improve policy robustness. |
Keywords
» Artificial intelligence » Regularization » Reinforcement learning
