{
  "id": "2502.03752",
  "version": "v1",
  "published": "2025-02-06T03:28:45.000Z",
  "updated": "2025-02-06T03:28:45.000Z",
  "title": "PRISM: A Robust Framework for Skill-based Meta-Reinforcement Learning with Noisy Demonstrations",
  "authors": [
    "Sanghyeon Lee",
    "Sangjun Bae",
    "Yisak Park",
    "Seungyul Han"
  ],
  "comment": "8 pages main, 19 pages appendix with reference. Submitted to ICML 2025",
  "categories": [
    "cs.LG",
    "cs.AI"
  ],
  "abstract": "Meta-reinforcement learning (Meta-RL) facilitates rapid adaptation to unseen tasks but faces challenges in long-horizon environments. Skill-based approaches tackle this by decomposing state-action sequences into reusable skills and employing hierarchical decision-making. However, these methods are highly susceptible to noisy offline demonstrations, resulting in unstable skill learning and degraded performance. To overcome this, we propose Prioritized Refinement for Skill-Based Meta-RL (PRISM), a robust framework that integrates exploration near noisy data to generate online trajectories and combines them with offline data. Through prioritization, PRISM extracts high-quality data to learn task-relevant skills effectively. By addressing the impact of noise, our method ensures stable skill learning and achieves superior performance in long-horizon tasks, even with noisy and sub-optimal data.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2025-02-06T03:28:45.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "robust framework",
      "skill-based meta-reinforcement learning",
      "noisy demonstrations",
      "ensures stable skill learning",
      "prism extracts high-quality data"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 8,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}