軌道優先：多様な方策を発見するためのカリキュラム
Trajectory First: A Curriculum for Discovering Diverse Policies
スプライン軌道を事前知識として多様な行動を生成し、その後反応的な方策に蒸留する2段階カリキュラムで、ロボット操作における多様性と性能を向上させる。
詳しい要約
1. どんなもの？
2. 先行研究と比べてどこがすごい？
3. 技術・手法の肝は？
4. どうやって有効だと検証した？
5. 議論はある？
6. 次に読むべき論文は？
※ AIが要旨から生成した要約です。正確性は原文をご確認ください。
著者: Cornelius V. Braun, Sayantan Auddy, Marc Toussaint
分類: cs.LG, cs.RO
原文アブストラクト
Being able to solve a task in diverse ways makes agents more robust to task variations and less prone to local optima. In this context, constrained diversity optimization has become a useful reinforcement learning (RL) framework for training a set of diverse agents in parallel. However, existing constrained-diversity RL methods often under-explore in complex tasks such as robot manipulation, resulting in limited behavioral diversity. We address this with a two-stage curriculum that introduces a spline-based trajectory prior as an inductive bias to produce diverse, high-reward behaviors in an initial stage, and then distills these behaviors into reactive, step-wise policies in a second stage. In our empirical evaluation, we provide novel insights into challenges of diversity-targeted training and show that our curriculum increases the diversity of learned skills while maintaining high task performance.
