AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

CATAL: Causally Disentangled Task Representation Learning for Offline Meta-Reinforcement Learning

Shan Cong, Chao Yu, Xiangyuan Lan

AAAI 2026 · Reinforcement learning › Visual and representation RL · approach: Learned state and action representations · 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 CATAL's causal disentanglement components (independence, consistency, remapping) to the base COMRL framework (ablation)CATAL · Hopper (ID)average return: 243.6 -> 289.9Improved
add CATAL's three components to the base framework (ablation)CATAL · Walker (OOD)average return: 150.3 -> 256.3Improved
add the consistency constraint (ablation)CATAL · Half-Cheetah (OOD)average return: -213.5 -> -184.2Improved
add the independence constraint (ablation)CATAL · Half-Cheetah (OOD)average return: -251.5 -> -184.2Improved
add the independence constraint (ablation)CATAL · Hopper (ID)average return: 196.6 -> 289.9Improved
replace CORRO with CATAL representationsBRAC meta-policy · Ant (OOD)average return: 136.4 -> 164.7Improved
replace CSRO with CATAL representationsBRAC meta-policy · Walker (OOD)average return: 230.3 -> 256.3Improved

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