Research map / Vision-language and multimodal / Robot action and navigation
RoboClaw: An Agentic Framework for Scalable Long-Horizon Robotic Tasks
Ruiying Li, Yunlang Zhou, Yuyao Zhu, Kylin Chen, Sukai Wang, Kongtao Hu, Minhui Yu, Bowen Jiang, Jiayao Ma, Zhan Su, Yongjian Shen, Yang Yang and 4 more
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add more rollout iterations (1 to 5, 50 samples each) of EAP data (ablation) | RoboClaw forward policy · Forward policy, Body Lotion | success rate: 21/50 -> 43/50 | Improved |
| add more rollout iterations (1 to 5, 50 samples each) of EAP data (ablation) | RoboClaw forward policy · Forward policy, Primer | success rate: 23/50 -> 40/50 | Improved |
| add more rollout iterations (1 to 5, 50 samples each) of EAP data (ablation) | RoboClaw forward policy · Forward policy, Lipstick | success rate: 2/50 -> 23/50 | Improved |
| add more rollout iterations (1 to 5, 50 samples each) of EAP data (ablation) | RoboClaw forward policy · Forward policy, Tissue Wipe | success rate: 11/50 -> 26/50 | Improved |
| use EAP autonomous collection instead of manual demonstrations and resets | RoboClaw · Data collection (4 scenes) | human time: manual baseline requires 2.16x more human time | Improved |
| use EAP autonomous collection instead of manual | RoboClaw · Data collection rollouts | human intervention: manual baseline requires 8.04x more human intervention | Improved |
| use RoboClaw agent monitoring and recovery vs pi0.5 baseline and product-of-subtask baseline | RoboClaw · Vanity table organization | success rate: 25% higher success rate | Improved |
| add human intervention requests on abnormal situations | RoboClaw · Human burden | human burden: reduces human burden by 53.7% | Improved |
| add learned recovery policies from rollouts to policy library | RoboClaw · Long-horizon execution | success rate | Improved |
Closest papers by meaning
- RoboCerebra: A Large-scale Benchmark for Long-horizon Robotic Manipulation Evaluation (NeurIPS 2025 · similarity 0.92)
- CaP-X: A Framework for Benchmarking and Improving Coding Agents for Robot Manipulation (ICML 2026 · similarity 0.89)
- ElegantVLA: Learning When to Think for Efficient Vision-Language-Action Models (NeurIPS 2026 · similarity 0.88)
- Plan-Seq-Learn: Language Model Guided RL for Solving Long Horizon Robotics Tasks (ICLR 2024 · similarity 0.88)
- LaST-R1: Reinforcing Robotic Manipulation via Adaptive Physical Latent Reasoning (NeurIPS 2026 · similarity 0.88)
- RoboMamba: Efficient Vision-Language-Action Model for Robotic Reasoning and Manipulation (NeurIPS 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
