2 papers
cs.CV2026
Reconstruction or Semantics? What Makes a Latent Space Useful for Robotic World Models
Nilaksh, Saurav Jha, Artem Zholus +1
World model-based policy evaluation is a practical proxy for testing real-world robot control by rolling out candidate actions in action-conditioned video diffusion models. As thes…
cs.LG2026
Squeezing More from the Stream : Learning Representation Online for Streaming Reinforcement Learning
Nilaksh, Antoine Clavaud, Mathieu Reymond +2
In streaming Reinforcement Learning (RL), transitions are observed and discarded immediately after a single update. While this minimizes resource usage for on-device applications,…