Freeze, Share, Shrink: Rethinking the Action Backbone in Diffusion Policies

arXiv:2511.12101 · cs.RO, cs.AI, cs.LG · Submitted 2025-11-15 · Read on arXiv

cs.RO, cs.AI, cs.LG

Submitted: 2025-11-15

Updated: 2026-09-12

Comments: RSS2026-Diff4RL

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Decoupling — 12× in this paper · explained in 2 episodes
• Transformer — 12× in this paper · explained in 2 episodes
• Ablation — 11× in this paper · explained in 1 episode
• Benchmark — 7× in this paper · explained in 2 episodes
• Diffusion Model — 6× in this paper · explained in 8 episodes
• Diffusion Models — 6× in this paper · explained in 6 episodes
• Foundation Model — 5× in this paper · explained in 3 episodes
• VLA Models — 4× in this paper · explained in 3 episodes
• Adaptation — 3× in this paper · explained in 1 episode
• Downstream Tasks — 3× in this paper · explained in 2 episodes
• Flow Matching — 2× in this paper · explained in 7 episodes
• Latent Diffusion Model — 2× in this paper · explained in 2 episodes
• Latent Diffusion Models — 2× in this paper · explained in 2 episodes
• Adaptive Reasoning — 1× in this paper · explained in 2 episodes
• Affordance — 1× in this paper · explained in 1 episode
• Benchmarking — 1× in this paper · explained in 3 episodes
• Core — 1× in this paper · explained in 2 episodes
• Distribution Shift — 1× in this paper · explained in 5 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Generalization — 1× in this paper · explained in 13 episodes
• Gradient Flow — 1× in this paper · explained in 2 episodes
• Robustness — 1× in this paper · explained in 5 episodes
• Shared Self-Attn — 1× in this paper
• Temporal Correlations — 1× in this paper · explained in 2 episodes
• Two-stage training — 1× in this paper · explained in 2 episodes
• World Model — 1× in this paper · explained in 7 episodes

Sources

• $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control
• CogACT: A Foundational Vision-Language-Action Model for Synergizing Cognition and Action in Robotic Manipulation
• GR00T N1: An Open Foundation Model for Generalist Humanoid Robots
• DexVLA: Vision-Language Model with Plug-In Diffusion Expert for General Robot Control
• Knowledge Insulating Vision-Language-Action Models: Train Fast, Run Fast, Generalize Better
• RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control
• OpenVLA: An Open-Source Vision-Language-Action Model
• $\pi_{0.5}$: a Vision-Language-Action Model with Open-World Generalization
• OneTwoVLA: A Unified Vision-Language-Action Model with Adaptive Reasoning
• Interleave-VLA: Enhancing Robot Manipulation with Interleaved Image-Text Instructions
• Evo-0: Vision-Language-Action Model with Implicit Spatial Understanding
• ACoT-VLA: Action Chain-of-Thought for Vision-Language-Action Models
• VLAW: Iterative Co-Improvement of Vision-Language-Action Policy and World Model
• SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics
• TinyVLA: Towards Fast, Data-Efficient Vision-Language-Action Models for Robotic Manipulation
• $\pi^{*}_{0.6}$: a VLA That Learns From Experience
• Gemma 3 Technical Report
• PointVLA: Injecting the 3D World into Vision-Language-Action Models
• GR-3 Technical Report
• HiMoE-VLA: Hierarchical Mixture-of-Experts for Generalist Vision-Language-Action Policies

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
• HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
• APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies
• Fine-tuning is Not Enough: A Parallel Framework for Collaborative Imitation and Reinforcement Learning in End-to-end Autonomous Driving