Explicitly capturing demonstrator multi-modality for distillation into diffusion policies
cs.RO
Submitted: 2026-11-07
Updated: 2026-11-07
Terminology
Sources
- $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control
- Data-Asymmetric Latent Imagination and Reranking for 3D Robotic Imitation Learning
- Posterior Behavioral Cloning: Pretraining BC Policies for Efficient RL Finetuning
- TMRL: Diffusion Timestep-Modulated Pretraining Enables Exploration for Efficient Policy Finetuning
- From Flow to One Step: Real-Time Multi-Modal Trajectory Policies via Implicit Maximum Likelihood Estimation-based Distribution Distillation
- Memorization and Regularization in Generative Diffusion Models
- Provable Separations between Memorization and Generalization in Diffusion Models
- On Memorization in Diffusion Models
- Demystifying Diffusion Policies: Action Memorization and Simple Lookup Table Alternatives
- SIME: Enhancing Policy Self-Improvement with Modal-level Exploration
- SOE: Sample-Efficient Robot Policy Self-Improvement via On-Manifold Exploration
- Scaling Relationship on Learning Mathematical Reasoning with Large Language Models
- Escaping the Mode Lottery: Multi-Response Training Improves Language Model Generalization
- A Probabilistic U-Net for Segmentation of Ambiguous Images
- Extracting Training Data from Diffusion 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