V-JEPA 2 Action-Conditioned & Robotics

Community

Plan with action-conditioned robotics (V-JEPA 2)

Authorsovr610
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Automates action-conditioned robotics planning and control.

Core Features & Use Cases

  • Interleaves action/state/extrinsics conditioning tokens with visual tokens to build a unified representation for learning and planning.
  • Enforces block-causal attention across frames for autoregressive prediction, enabling robust MPC in robotics.
  • Supports Action-Conditioned RoPE (AC-RoPE) differentiation between visual and conditioning tokens, plus CEM-based model-predictive control and DROID-style data loading.

Quick Start

Run a sample MPC planning session using the included synthetic world model to observe CEM convergence.

Dependency Matrix

Required Modules

numpypytesttorchopencv-pythonh5pyscipy

Components

scriptsreferencesassets

💻 Claude Code Installation

Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.

Please help me install this Skill:
Name: V-JEPA 2 Action-Conditioned & Robotics
Download link: https://github.com/sovr610/refffiy/archive/main.zip#v-jepa-2-action-conditioned-robotics

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.