OnlyAct: Extracting Action Representations from Foundation Models for World Models
cs.CV
Submitted: 2026-11-10
Updated: 2026-11-10
Code: https://github.com/google-research/robodesk
Terminology
Sources
- Cosmos World Foundation Model Platform for Physical AI
- AgiBot World Colosseo: A Large-scale Manipulation Platform for Scalable and Intelligent Embodied Systems
- V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning
- DeepMind Lab
- VFMF: World Modeling by Forecasting Vision Foundation Model Features
- UniVLA: Learning to Act Anywhere with Task-centric Latent Actions
- Leveraging Procedural Generation to Benchmark Reinforcement Learning
- ConLA: Contrastive Latent Action Learning from Human Videos for Robotic Manipulation
- DreamDojo: A Generalist Robot World Model from Large-Scale Human Videos
- Matrix-game 2.0: An open-source, real-time, and streaming interactive world model
- Classifier-Free Diffusion Guidance
- EgoDex: Learning Dexterous Manipulation from Large-Scale Egocentric Video
- Voyager: Long-Range and World-Consistent Video Diffusion for Explorable 3D Scene Generation
- Language-Driven Representation Learning for Robotics
- HunyuanVideo: A Systematic Framework For Large Video Generative Models
- Qwen3-VL-Embedding and Qwen3-VL-Reranker: A Unified Framework for State-of-the-Art Multimodal Retrieval and Ranking
- VIP: Towards Universal Visual Reward and Representation via Value-Implicit Pre-Training
- Yume: An Interactive World Generation Model
- ActionParty: Multi-Subject Action Binding in Generative Video Games
- Advancing Open-source World Models
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs