RecPo:通过重构程序性知识实现机器人组合式技能适应
Reconfiguring Procedural Knowledge for Compositional Robot Skill Adaptation
AI 导读
RecPo 框架通过任务无关的多个 procedure view 表示技能,组合已存储的技能转移序列来生成可执行技能序列,实现在稀疏任务信息下的行为重构。在扩展的 Franka Kitchen 长程任务基准上,RecPo 达到 91.9% 的有序完成率,每个 episode 的任务推理时间仅 1.0 ms。真机实验还表明,无需任务特定重训练,仅通过局部记忆编辑即可完成技能适应。
正文 · 原文
Abstract:General-purpose behavior models increasingly internalize broad procedural knowledge during training, making their adaptation to new goals and skill interfaces less a problem of learning entirely new behaviors than of reconfiguring existing competence. We study this capability as reconfigurability, the ability to realize new behaviors by cost-efficiently restructuring existing components and relations under sparse governing information, where task specifications provide high-level goals and procedure specifications expose a limited set of reusable factors, leaving the concrete bindings and compositions required for execution underdetermined. We propose RecPo, a framework that represents skills task-agnostically through multiple procedure views and materializes executable skill sequences by composing stored transitions to realize task-specific changes. RecPo maintains a procedure knowledge base of skill-induced transitions, with each view exposing a reusable factor and preserving local skill connectivity. At deployment, it interprets task specifications through the same views and resolves the resulting compositional queries over the stored relations to bind skills for execution. On long-horizon tasks in an extended Franka Kitchen benchmark, RecPo achieves 91.9% ordered completion with 1.0 ms of task inference per episode across varying procedure support and task-record sparsity. Real-robot studies further demonstrate adaptation through local memory edits without task-specific retraining.
| Comments: | Accepted at NeurIPS 2026 |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.22174 [cs.AI] |
| (or arXiv:2611.22174v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.22174 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Daehee Lee [view email]
[v1]
Tue, 17 Nov 2026 04:09:02 UTC (11,179 KB)
来源:arXiv cs.AI · arxiv.org