PrimitiveVLA: Learning Reusable Motion Primitives for Efficient and Generalizable Robotic Manipulation
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866916056261459968 |
|---|---|
| author | Li, Yutai Peng, Shaohui Guo, Jiaming Huang, Di Zhang, Zihao Guo, Yuxuan Gao, Yunkai Lan, Siming Li, Ling Hu, Xing Chen, Yunji |
| author_facet | Li, Yutai Peng, Shaohui Guo, Jiaming Huang, Di Zhang, Zihao Guo, Yuxuan Gao, Yunkai Lan, Siming Li, Ling Hu, Xing Chen, Yunji |
| contents | Vision-Language-Action (VLA) models offer a promising paradigm for generalist robotic policies, yet their adaptation is hindered by data inefficiency and poor generalization. We argue that these bottlenecks stem from the prevailing Direct Instruction-to-Control Mapping, which forces models to memorize monolithic trajectories rather than reusable motion patterns, i.e., primitives. We propose PrimitiveVLA, a framework that shifts this paradigm toward a Primitive-Centric Disassemble & Assemble paradigm. Supported by a shared Multimodal Canonical Representation (MCR), PrimitiveVLA unifies two phases: (1) Fine-tuning-phase Disassembly, which uses an automated pipeline to disassemble demonstrations into reusable primitives; and (2) Inference-phase Assembly, which employs a VLM-based planner and an LLM-generated switch module for robust closed-loop execution. By disassembling tasks into reusable primitives, PrimitiveVLA enables VLA models to learn invariant motion patterns instead of task-specific trajectories. Extensive experiments show that our framework improves data efficiency and achieves superior zero-shot generalization across unseen and long-horizon tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_28634 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | PrimitiveVLA: Learning Reusable Motion Primitives for Efficient and Generalizable Robotic Manipulation Li, Yutai Peng, Shaohui Guo, Jiaming Huang, Di Zhang, Zihao Guo, Yuxuan Gao, Yunkai Lan, Siming Li, Ling Hu, Xing Chen, Yunji Robotics Vision-Language-Action (VLA) models offer a promising paradigm for generalist robotic policies, yet their adaptation is hindered by data inefficiency and poor generalization. We argue that these bottlenecks stem from the prevailing Direct Instruction-to-Control Mapping, which forces models to memorize monolithic trajectories rather than reusable motion patterns, i.e., primitives. We propose PrimitiveVLA, a framework that shifts this paradigm toward a Primitive-Centric Disassemble & Assemble paradigm. Supported by a shared Multimodal Canonical Representation (MCR), PrimitiveVLA unifies two phases: (1) Fine-tuning-phase Disassembly, which uses an automated pipeline to disassemble demonstrations into reusable primitives; and (2) Inference-phase Assembly, which employs a VLM-based planner and an LLM-generated switch module for robust closed-loop execution. By disassembling tasks into reusable primitives, PrimitiveVLA enables VLA models to learn invariant motion patterns instead of task-specific trajectories. Extensive experiments show that our framework improves data efficiency and achieves superior zero-shot generalization across unseen and long-horizon tasks. |
| title | PrimitiveVLA: Learning Reusable Motion Primitives for Efficient and Generalizable Robotic Manipulation |
| topic | Robotics |
| url | https://arxiv.org/abs/2605.28634 |