arXiv:2606.04130 · v2

CLAW: Continuous Latent Action World Models via Adversarial Regularization

首次提交 2026-06-03 02:40

版本更新 2026-11-06 05:05

雷达观测 2026-11-09 10:16

摘要

We introduce CLAW, a fully end-to-end self-supervised framework for learning a world model jointly with continuous latent action representations directly from action-free videos. Our approach leverages adversarial latent regularization to prevent the latent from encoding future observations, so that it captures the structure of the action. The diffusion-based world model then models environment dynamics as a distribution over future observations rather than as a single deterministic rollout. Both components are learned from observation sequences alone without any action annotations. By simultaneously training the Latent Action Model and world model, CLAW learns to reason about how inferred actions induce environment transitions from visual observations alone. We show that the resulting latent action world model supports both imitation learning from observation and goal-directed planning. In imitation learning, latent actions extracted from raw videos enable behavior cloning. For planning, CLAW generates sequences of latent actions and maps them to executable actions to reach desired goals. Extensive experiments across diverse tasks and embodiments demonstrate that CLAW produces semantically meaningful latent action representations, supports effective action transfer, and enables planning and imitation from observation, outperforming existing methods.

作者备注：9 pages, 27 pages of supplementary material