Research map / Learning theory and algorithms / Combinatorial and search optimization
CABTO: Context-Aware Behavior Tree Grounding for Robot Manipulation
Yishuai Cai, Xinglin Chen, Yunxin Mao, Kun Hu, Minglong Li
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 planning context feedback (ablation) | CABTO, GPT-4o · 7 task sets | CSR: 50% -> 90.0% | Improved |
| add planning context feedback (ablation) | CABTO, GPT-3.5-Turbo · 7 task sets | ASR: 67.1% -> 74.8% | Improved |
| add planning context feedback (ablation) | CABTO, GPT-3.5-Turbo · Franka Blocks and Fetch Cook Meal | ASR: 70.0% -> 70.0% | No clear effect |
| add planning context feedback (ablation) | CABTO, GPT-4o · Dual-Franka Pour | ASR: 66.7% -> 100.0% | Improved |
| use GPT-4o instead of GPT-3.5-Turbo with planning context | CABTO · 7 task sets | CSR: 64.3% -> 90.0% | Improved |
| use GPT-4o instead of GPT-3.5-Turbo without planning context | CABTO · 7 task sets | ASR: 67.1% -> 66.2% | No clear effect |
| add execution contexts to VLM policy sampling (ablation) | Molmo+cuRobo+APIs · 5 action models | success rate: 44% -> 62% | Improved |
| use VLM hierarchical sampler instead of OpenVLA | policy sampling · 5 action models | success rate: 28% -> 62% | Improved |
| add environment feedback to action-model refinement (ablation) | cross-level refinement · 5 defective action models | success rate: 44% -> 74% | Improved |
| refine Put(obj, loc) with feedback (ablation) | cross-level refinement · Put delete-effect defect | success rate: 20% -> 40% | Mixed |
Closest papers by meaning
- MRBTP: Efficient Multi-Robot Behavior Tree Planning and Collaboration (AAAI 2025 · similarity 0.88)
- Instruction-Augmented Long-Horizon Planning: Embedding Grounding Mechanisms in Embodied Mobile Manipulation (AAAI 2025 · similarity 0.86)
- RoboCerebra: A Large-scale Benchmark for Long-horizon Robotic Manipulation Evaluation (NeurIPS 2025 · similarity 0.86)
- UniDomain: Pretraining a Unified PDDL Domain from Real-World Demonstrations for Generalizable Robot Task Planning (NeurIPS 2025 · similarity 0.85)
- Behavioral Exploration: Learning to Explore via In-Context Adaptation (ICML 2025 · similarity 0.85)
- Efficient Robotic Policy Learning via Latent Space Backward Planning (ICML 2025 · similarity 0.85)
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.
