AI Paper Atlas

Research map / Reinforcement learning / Policy gradient and actor-critic

Mind Dreamer: Untethering Imagination via Active Causal Intervention on Latent Manifolds

Shaojun Xu, Xiaoling Zhou, Yihan Lin, Yapeng Meng, Xinglong Ji, Luping Shi, Rong Zhao

ICML 2026 · Reinforcement learning › Policy gradient and actor-critic · approach: Regularized scalable deep value learning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
sample imagination start states from adversarial generator instead of replay bufferDreamerV3 · DMC vision (20 tasks)steps to 90% peak performance: 557.6k -> 334.7kImproved
sample imagination start states from adversarial generatorDreamerV3 · DMC vision (20 tasks)average return: 780.3 -> 831.1Improved
add manifold constraint Lmf (ablation)Mind Dreamer · DMCreturnImproved
add Relay Uncertainty Function (epistemic) (ablation)Mind Dreamer · DMCreturnImproved
add Relay Value Function (pragmatic) (ablation)Mind Dreamer · DMCreturnImproved
use generator-based jumps on Three-Ring ManifoldDreamerV3 · Three-Ring Manifoldfirst hitting time: 4.2x accelerationImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy