arXiv：cs.LG（机器学习，全量分类）· Mohammad Amin Abbasfar, Farbod Azimmohseni, Mohammad Hossein Rohban·· 14 小时前AI 评分32
用目标条件化双模拟学习可迁移技能
Learning Transferable Skills using Goal-Conditioned Bisimulation
AI 导读
研究者提出一种目标条件化双模拟的无监督技能发现方法，通过学习满足函数等变性、保留环境局部时序结构的动作感知时序表示，让技能行为仅依赖直接影响其执行的状态特征子集，从而在不同布局间保持行为不变。实验表明，在给定环境中习得的技能可有效迁移到多种环境布局，解决下游任务，展现出较强的分布外泛化能力。
正文
Abstract:Unsupervised skill discovery has emerged as a promising approach for leveraging reward-free datasets to pretrain general-purpose policies. However, current skill discovery methods either require access to expert data or exhibit limited generalization, failing to transfer effectively to previously unseen layouts. A key challenge is to learn representations that capture the temporal structure of the environment while remaining robust to variations across layouts. To address this issue, we present an objective for learning action-aware temporal representations that satisfy the functional equivariance property while preserving the local temporal structure of the environment. Building upon this embedding, we further propose unsupervised skill discovery using bisimulation, which learns transferable skills by conditioning the behavior of skills exclusively on the subset of state features that directly affect their execution. This enforces invariant behavior across different layouts, enabling skills to transfer effectively to other configurations. Finally, through comprehensive empirical evaluations, we show that skills learned in a given environment can be effectively applied to solve downstream tasks in various environment layouts, demonstrating strong out-of-distribution generalization.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Robotics (cs.RO) |
| Cite as: | arXiv:2610.00676 [cs.LG] |
| (or arXiv:2610.00676v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.00676 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Mohammad Amin Abbasfar [view email]
[v1]
Wed, 30 Sep 2026 20:13:26 UTC (1,065 KB)
来源：arXiv：cs.LG（机器学习，全量分类） · arxiv.org