AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Model-Based Reinforcement Learning via Imagination with Derived Memory

Yao Mu, Yuzheng Zhuang, Bin Wang, Guangxiang Zhu, Wulong Liu, Jianyu Chen, Ping Luo, Shengbo Eben Li, Chongjie Zhang, Jianye HAO

NeurIPS 2021 · Reinforcement learning › Visual and representation RL · approach: Latent-space world models · 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
add derived memory, R2S2 decoding constraint and evaluator (IDM vs Dreamer)Dreamer · Walker-Walk, image uncertainty, 500Kaverage return: 876.3 ± 48 -> 914.9 ± 61Improved
add IDM to DreamerDreamer · Hopper-Stand, image uncertainty, 500Kaverage return: 430.2 ± 128 -> 841.2 ± 54Improved
add IDM to DreamerDreamer · Walker-Walk, no uncertainty, 100Kaverage return: 277 ± 12 -> 630 ± 113Improved
IDM vs model-free DrQDrQ · Cheetah-Run, no uncertainty, 500Kaverage return: 660 ± 96 vs 631 ± 103Hurt
IDM vs DrQ under test-time image rotationDrQ · Walker-Walk, image uncertainty, 500Kaverage return: 151.9 ± 23 -> 914.9 ± 61Improved
remove trajectory evaluator (IDM w/o EVA) (ablation)IDM · Hopper-Stand, image uncertaintyaverage returnHurt
remove trajectory evaluator (IDM w/o EVA) (ablation)IDM · Walker-Walk, image uncertaintyaverage returnNo clear effect
remove derived memory, keep constraint and evaluator (IDM w/o DM) (ablation)IDM · Walker-Walk, image uncertainty, 100Kimprovement over Dreamer: 26.6% vs 41.6%Hurt
remove decoding constraint and derived memory, keep evaluator only (ablation)IDM · DMControl, image uncertaintyaverage returnNo clear effect
double Dreamer batch size (Doubled-Dreamer) (ablation)Dreamer · DMControl, image uncertaintyaverage returnNo clear effect
derived memory contribution on stable Walker-Walk (ablation)IDM · Walker-Walkaverage returnImproved
evaluator contribution on unstable Hopper-Stand (ablation)IDM · Hopper-Standaverage returnImproved

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