Compounding Physical RSI: Turning Execution Failure into Inherited Capability
cs.RO
Submitted: 2026-11-04
Updated: 2026-11-04
Code: https://github.com/NVIDIA/Isaac-GR00T
Terminology
Sources
- Spatial Forcing: Implicit Spatial Representation Alignment for Vision-language-action Model
- Xiaomi-Robotics-1: Scaling Vision-Language-Action Models with over 100K Hours of Real-World Trajectories
- The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement
- Who Improves the Robot? Toward Physical Recursive Self-Improvement
- Toward Recursive Self-Improvement in Physical AI
- CycleVLA: Proactive Self-Correcting Vision-Language-Action Models via Subtask Backtracking and Minimum Bayes Risk Decoding
- Self-Adaptive VLA for Robust Robot Deployment
- ME-Brain-1.0: Memory, Cognition and Action for Evolving Embodied Intelligence
- $\pi^{*}_{0.6}$: a VLA That Learns From Experience
- EmbodiRSI: Recursive Self-Improvement for Data-Efficient Robot Adaptation
- Offline Reinforcement Learning: Tutorial, Review, and Perspectives on Open Problems
- Behavior Regularized Offline Reinforcement Learning
- RoboDojo: A Unified Sim-and-Real Benchmark for Comprehensive Evaluation of Generalist Robot Manipulation Policies
- XPolicyLab: A Unified Standard and Open Ecosystem for Robot Policy Evaluation and Deployment
- G0.5: One Autoregressive Stream for Robot Reasoning and Action
- Hy-Embodied-0.5-VLA: From Vision-Language-Action Models to a Real-World Robot Learning Stack
- SimpleMemVLA: A Simple but Effective Native-Video Memory for Vision-Language-Action Models
- InternVLA-A1.5: Unifying Understanding, Latent Foresight, and Action for Compositional Generalization
- Beyond Data Scaling: Representation-Centric Continued Pre-training for Vision-Language-Action Models
- Unified 4D World Action Modeling from Video Priors with Asynchronous Denoising
Related papers
- Rethinking Actions: Continuity-Aware Priors for World-Model Planning
- MOSAIC: Modular Obstacle and Spatial-Aware Framework for Manipulation In Cluttered Scenes
- SAGAS: Semantic-Aware Graph-Assisted Stitching for Offline Temporal Logic Planning
- DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
- LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
- CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation