VLA-JEPA: Enhancing Vision-Language-Action Model with Latent World Model

Zero-to-mastery JEPA study guide: 245 cells, 14 GIFs, interactive notebooks covering JEPA, V-JEPA 2, VLA models, world models, and ICRA research

This repository contains latest research papers relating to VLA models and World Models like JEPA in the domain of embodied AI.

Obsidian knowledge vault covering physical intelligence research: JEPA, world models, VLA, video planning — 34 papers, 64 metrics, 78 datasets with inline-linked tables and embedded PDFs

Personal learning space for embodied AI & robotics — covering the full stack from robot hardware and perception (VLM) to task planning (LLM + CoT), vision-language-action models (VLA), world models (Dreamer, JEPA), and a runnable browser demo.