Few-Shot Residual Adaptation of Pretrained Latent World Models
cs.RO
Submitted: 2026-11-10
Updated: 2026-11-10
Terminology
Sources
- V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning
- Mastering Diverse Domains through World Models
- Self-Supervised Policy Adaptation during Deployment
- World Model for Robot Learning: A Comprehensive Survey
- LoRA: Low-Rank Adaptation of Large Language Models
- DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset
- RMA: Rapid Motor Adaptation for Legged Robots
- Adapting World Models with Latent-State Dynamics Residuals
- Simulation Distillation: Pretraining World Models in Simulation for Rapid Real-World Adaptation
- PIN-WM: Learning Physics-INformed World Models for Non-Prehensile Manipulation
- DINOv3
- Tent: Fully Test-time Adaptation by Entropy Minimization
- AdaJEPA: An Adaptive Latent World Model
- RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
- AdaWorldPolicy: World-Model-Driven Diffusion Policy with Online Adaptive Learning for Robotic Manipulation
- DINO-WM: World Models on Pre-trained Visual Features enable Zero-shot Planning
Related papers
- Rethinking Actions: Continuity-Aware Priors for World-Model Planning
- MOSAIC: Modular Obstacle and Spatial-Aware Framework for Manipulation In Cluttered Scenes
- SAGAS: Semantic-Aware Graph-Assisted Stitching for Offline Temporal Logic Planning
- DexSeed: Expanding Human Hand-Object Interactions into Robot Demonstrations for Dexterous Policy Learning
- LaST0.5: A Generalist World Action Model with Interaction-Centric Latent Reasoning
- CIALLO: Learning Shared Intent Representations for Cross-Embodiment Manipulation