AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Factored Adaptation for Non-Stationary Reinforcement Learning

Fan Feng, Biwei Huang, Kun Zhang, Sara Magliacane

NeurIPS 2022 · Reinforcement learning › Visual and representation RL · approach: In-context and adaptive RL · 6 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
remove factored representation (binary masks/structure) (ablation)FANS-RL · Half-Cheetah ablationaverage returnImproved
add state prediction (ablation)FANS-RL · Half-Cheetah ablationaverage returnImproved
add reward prediction (ablation)FANS-RL · Half-Cheetah with non-stationary rewardaverage returnImproved
add smoothness loss (ablation)FANS-RL · Half-Cheetah within-episode changesaverage returnImproved
use disentangled CF inference networks (ablation)FANS-RL · Half-Cheetah dynamics and reward changesaverage returnImproved
use FANS-RL instead of ZeUS, Meld, CaDM, Hyper-DynamicsFANS-RL · Sawyer-Peg raw pixelsaverage return: 12.45 -> 18.01Improved
increase non-stationarity degreeFANS-RL · Half-Cheetahaverage returnImproved
use fewer latent dimensionsFANS-RL · Half-Cheetahaverage returnImproved
add multiple concurrent change factors (1D to 2D+1R)FANS-RL · Half-Cheetahaverage returnImproved
vary functional form (piecewise linear, damped, sine)FANS-RL · Half-Cheetahaverage 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