ODE Policy: Robot Learning via Action Dynamics, Not Actions
cs.RO
Submitted: 2026-11-01
Updated: 2026-11-01
Terminology
Sources
- GR00T N1: An Open Foundation Model for Generalist Humanoid Robots
- $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control
- Where Should Action Generation Begin? A Learnable Source Prior for Generative Robot Policies
- OMP: One-step Meanflow Policy with Directional Alignment
- LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Models
- MARS Policy: Multimodality Only When It Matters
- Why Does Action Chunking Improve Behavioral Cloning Performance in Robotic Control?
- One-Step Flow Policy: Self-Distillation for Fast Visuomotor Policies
- SuperMimic: Scaling Humanoid Whole-Body Control with One Differentiable World Model
- MOSS: Mode-Oriented Source Shaping for Robot Flow Matching Policies
- $\pi_{0.5}$: a Vision-Language-Action Model with Open-World Generalization
- Warm Starts Accelerate Conditional Diffusion
- Proximal Policy Optimization Algorithms
- End-to-End Stable Imitation Learning via Autonomous Neural Dynamic Policies
- Chronos Policy: Learning Robot Policies on Physical Time
- ImageWAM: Do World Action Models Really Need Video Generation, or Just Image Editing?
- APEX: Adaptive Policy Execution for Precise Manipulation
- robosuite: A Modular Simulation Framework and Benchmark for Robot Learning
- Flow Matching for Robot Foundation Models
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