MobiAgent: Dual-Loop Recursive Policy Self-Improvement for Long-Horizon Mobile Manipulation
cs.RO, cs.AI
Submitted: 2026-10-02
Updated: 2026-10-02
Comments: Accepted at the Conference on Robot Learning (CoRL) 2026. Project page: https://kaiknower.github.io/mobiagent
Code: https://github.com/kaiknower/MobiAgent
Project page: https://kaiknower.github.io/mobiagent
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Sources
- N2M: Bridging Navigation and Manipulation by Learning Pose Preference from Rollout
- VLA^2: Empowering Vision-Language-Action Models with an Agentic Framework for Unseen Concept Manipulation
- RoboClaw: An Agentic Framework for Scalable Long-Horizon Robotic Tasks
- Task adaptation of Vision-Language-Action model: 1st Place Solution for the 2025 BEHAVIOR Challenge
- CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation
- Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success
- Sci-VLA: Agentic VLA Inference Plugin for Long-Horizon Tasks in Scientific Experiments
- Goal2Skill: Long-Horizon Manipulation with Adaptive Planning and Reflection
- CaP-X: A Framework for Benchmarking and Improving Coding Agents for Robot Manipulation
- When to Trust Imagination: Adaptive Action Execution for World Action Models
- MG-Nav: Dual-Scale Visual Navigation via Sparse Spatial Memory
- Towards Human-level Intelligence via Human-like Whole-Body Manipulation
- LoRA: Low-Rank Adaptation of Large Language Models
- PicoPose: Progressive Pixel-to-Pixel Correspondence Learning for Novel Object Pose Estimation
Related papers
- FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
- MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
- RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
- ARC: A Reasoning Recipe for Robot Foundation Models
- HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
- APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies