Hugging Face Daily PapersZhongbo Zhang, Zaibin Zhang, Yifan WangSep 17, 20261 min readpaperadvanced

Learning Foresight without Explicit Trajectories for 3D Diffusion Policies

Summary

This paper introduces Movement Trend Guidance (MTG), a method to provide foresight to 3D diffusion policies for robotic manipulation without explicit trajectory planning. MTG learns a compact latent representation of interaction evolution, significantly improving performance on various benchmarks with minimal parameter overhead.

3D diffusion policies often lack explicit foresight for anticipating interaction evolution.
Movement Trend Guidance (MTG) learns a compact latent representation of interaction trends from observation history.
This latent is supervised by sparse future gripper states during training and used as global conditioning during inference.
MTG adds only 3.52% more parameters to DP3 but substantially improves performance on RoboTwin2.0, LIBERO-40, and DexArt.

Robotics engineers working with diffusion policies for manipulation tasks should care, as this method offers a simple yet effective way to enhance policy performance by incorporating future awareness.

7/10

machine learning reinforcement learning computer vision robotics diffusion models

Related reading

From Pretraining to Proficiency: Real-World Subtask RL for Long-Horizon Manipulation with Minimal Human Intervention arxiv.org
The paper presents PARTS, a framework that augments a frozen pretrained robot policy with RL‑learned residuals on selected bottleneck subtasks, using local success rewards and minimal human resets. In real‑world bimanual and single‑arm tasks, PARTS more than doubles success rates with only minutes of robot rollouts, outperforming prior fine‑tuning methods.
Hugging Face Daily Papersarxiv.orgSichang Su, Benjamin Yang, Zhiyun Deng4d1 minpaper
HarnessVLN: Unifying Training-Free Embodied Navigation through an Agent Harness arxiv.org
HarnessVLN introduces a zero‑shot, training‑free embodied navigation framework that wraps a multimodal LLM in an "Agent Harness" – a tool‑based protocol that validates planner actions against spatial evidence, tracks progress with hierarchical event memory, and maintains a persistent spatiotemporal graph for recovery. The system works for instruction‑following and object‑goal tasks, achieving 60.…
Hugging Face Daily Papersarxiv.orgYang Chen, Lirong Che, Zhenyu Huang8d1 minpaper
Breaking the Vision-Action Shortcut: Latent Interface Training for Generalizable Robotics Foundation Models arxiv.org
Latent Interface Training (LIT) first learns a goal‑conditioned action prior without visual input, then adds a pose‑supervised latent interface as the only visual conditioning path. Applied to several vision‑language‑action models, LIT cuts vision‑action shortcuts and lifts LIBERO‑Plus success by 3.9–10.7 points and real‑world task success by 13.3–16.7 points under distribution shifts.
Hugging Face Daily Papersarxiv.orgJianman Lin, Shailesh Shailesh, Zhongyi Luo11d1 minpaper
GAVEL: Graph World Models for Verified and Efficient Long-Horizon LLM Task Planning arxiv.org
GAVEL augments LLM‑driven robot planners with an explicit graph world model that verifies actions, repairs violations, and reasons over belief distributions, boosting single‑task success from 41 % to 92 % and multi‑task success from 20 % to 93 % on BEHAVIOR‑1K.
Hugging Face Daily Papersarxiv.orgRuiyang Wang, Hao-Lun Hsu, Swarajh Mehta6d1 minpaper
In-Context Robot Learning with VLM Agents arxiv.org
GPT‑Policy is a framework that lets a large vision‑language model (e.g. GPT‑6 Astra) perform in‑context robot learning: a context compiler extracts visual transitions from demos, the VLM proposes tool actions, and a constrained controller verifies and executes them. Real‑robot experiments show that raw video demos improve success rates even without explicit action labels, and that providing align…
Hugging Face Daily Papersarxiv.orgDongzhou Cheng, Taoran Yi, Ye Fang6d1 minpaper
ActionPiece: Rethinking Action Tokenization for Autoregressive Vision-Language-Action Models arxiv.org
Action tokenizers for autoregressive VLA models often fail to preserve subtle action adjustments, despite good pointwise reconstruction. This paper introduces Physical Rank Consistency (PRC) to measure relational fidelity and ActionPiece, a new tokenizer that uses joint supervision to preserve these physical relationships, significantly improving policy success on robotics benchmarks.
Hugging Face Daily Papersarxiv.orgShijie Lian, Bin Yu, Zhaolong Shen6d1 minpaper

Learning Foresight without Explicit Trajectories for 3D Diffusion Policies

Summary

Related reading

From Pretraining to Proficiency: Real-World Subtask RL for Long-Horizon Manipulation with Minimal Human Interventionarxiv.org

HarnessVLN: Unifying Training-Free Embodied Navigation through an Agent Harnessarxiv.org

Breaking the Vision-Action Shortcut: Latent Interface Training for Generalizable Robotics Foundation Modelsarxiv.org

GAVEL: Graph World Models for Verified and Efficient Long-Horizon LLM Task Planningarxiv.org

In-Context Robot Learning with VLM Agentsarxiv.org

ActionPiece: Rethinking Action Tokenization for Autoregressive Vision-Language-Action Modelsarxiv.org

From Pretraining to Proficiency: Real-World Subtask RL for Long-Horizon Manipulation with Minimal Human Intervention arxiv.org

HarnessVLN: Unifying Training-Free Embodied Navigation through an Agent Harness arxiv.org

Breaking the Vision-Action Shortcut: Latent Interface Training for Generalizable Robotics Foundation Models arxiv.org

GAVEL: Graph World Models for Verified and Efficient Long-Horizon LLM Task Planning arxiv.org

In-Context Robot Learning with VLM Agents arxiv.org

ActionPiece: Rethinking Action Tokenization for Autoregressive Vision-Language-Action Models arxiv.org