arXiv Machine Learning
Dream to Remember: Continual Steering of Generalist Robot Policies with World Models
Read the original on arXiv Machine Learning →The Flow has not summarised this story yet — read it at arXiv Machine Learning.
Related stories
arXiv AI
Jul 23
The World Model Remembers, the Actor Forgets: Dream Rehearsal for Continual Model-Based RL
arXiv:2607. 19749v1 Announce Type: cross Abstract: Model-based reinforcement-learning agents of the DreamerV3 family forget catastrophically when trained on task sequences, even when an unbounded replay buffer preserves every earlier experience.
arXiv Machine Learning
Sep 21
Benchmarking World Models for Continual Learning on Compositional Tasks
The paper introduces a compositional continual learning benchmark for world models in robot manipulation, designed to isolate knowledge reuse from learning speed and capacity. Tasks are curated to combine previously seen action and perception components, allowing analysis of how different modalities affect reuse. Experiments show that modular world models better balance reuse and forgetting than conventional methods, yet none fully solve the challenge, highlighting the need for models explicitly built to reuse knowledge without forgetting.
arXiv AI
Oct 5
RIFAR: Reliability and Forgetting-Aware Replay for Continual Robot Learning
RIFAR is a new continual learning method for robots that uses reliability screening and drift-aware replay to mitigate forgetting while keeping storage low. It reconstructs past trajectories from short demonstration prefixes and employs a frozen inverse-dynamics model to verify action‑visual consistency. In experiments on LIBERO suites and real‑world tasks, RIFAR outperforms previous generative replay approaches, achieving high performance with only a small fraction of stored steps.
arXiv Computer Vision
Oct 1
The Planning Limits of Latent World Models
arXiv:2609.39235v1 Announce Type: cross Abstract: World models offer a promising way to help robots understand how the physical world evolves and plan complex behaviours through imagination. Yet exis...
arXiv Machine Learning
5d ago
Learning from Hindsight for VLA Reinforcement Learning
arXiv:2607.09042v2 Announce Type: replace Abstract: Reinforcement learning is increasingly used to fine-tune vision-language-action (VLA) models, but robot interaction is expensive and learning becom...
arXiv Computer Vision
Sep 22
Think Like a World Model, Act Like a VLA: Distilling World-Model Representations into Compact Robot Policies
The paper proposes a method to distill world‑model representations into compact Vision‑Language‑Action (VLA) policies. By adding a single feature‑alignment term during VLA training, a frozen world model’s internal features are cached and the student policy learns to match them, eliminating the need for a generative future‑rolling component. The resulting lightweight policy runs in 32 ms on an RTX 5090, achieving high performance on LIBERO and RoboCasa‑GR1, and transfers effectively to real robotic hardware.