Abstract:Compositional generalization is essential for reaching unseen goals under novel contextual variations in offline goal-conditioned reinforcement learning (GCRL), where a generalist goal-reaching agent must be learned from limited data. Most prior approaches pursue this via trajectory stitching over temporally contiguous segments, which limits composing behaviors across varying contexts. To overcome this limitation, we formalize analogy transduction as synthesizing new plans by composing task-endogenous analogies with given contexts and propose a novel analogy representation tailored for it. Grounded in our theory, this analogy representation captures what changes under optimal task execution, remains invariant to contextual variations, and is sufficient for optimal goal reaching. We further contend that generalization to unseen analogy-context pairs is a practical obstacle in analogy transduction, and introduce a new approach for offline GCRL that enables analogy transduction beyond seen pairs to unseen combinations. We empirically demonstrate the effectiveness of our approach on OGBench manipulation environments, substantially outperforming prior methods that do not perform analogy transduction. Project page: this https URL
| Comments: | ICML 2026 |
| Subjects: | Machine Learning (cs.LG) |
| Cite as: | arXiv:2605.20609 [cs.LG] |
| (or arXiv:2605.20609v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2605.20609 arXiv-issued DOI via DataCite (pending registration) |
From: Junseok Kim [view email]
[v1]
Wed, 20 May 2026 01:54:18 UTC (5,395 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。