AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

TaskLoom: Weaving Knowledge Across Tasks in World Models

Qingzhang Zeng, Peixi Peng, Hang Li, Luntong Li, Yonghong Tian

ICML 2026 · Reinforcement learning › Visual and representation RL · approach: Latent-space world models

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 gradient-based task grouping (one world model per group) instead of one shared world model for all tasks (ablation)Shared-Dreamer · Proprio Controlmean return: 280.3 -> 556.8Improved
naively train all tasks jointly in a single shared world model (ablation)DreamerV3 · Proprio Controlmean return: 456.1 -> 280.3Hurt
add inter-group shared knowledge (SK) base in Stage 2 (ablation)G-Baseline · Proprio Controlmean return: 556.8 -> 647.4Improved
add inter-group shared knowledge (SK) base in Stage 2 (ablation)G-Baseline · Proprio Controlmedian return: 533.4 -> 775.7Improved
keep group-specific dynamics predictor decoupled from the shared knowledge base instead of absorbing it (ablation)TaskLoom · Proprio Controlmean return: 604.5 -> 647.4Improved
keep group-specific dynamics predictor decoupled from the shared knowledge base instead of absorbing it (ablation)TaskLoom · Proprio Controlmedian return: 704.7 -> 775.7Improved
add task grouping without cross-group knowledge flow (ablation)Shared-Dreamer · Hopper Hop (Proprio)return: 3.9 -> 0.6Hurt
apply TaskLoom cross-task knowledge sharing instead of single-task DreamerV3DreamerV3 · Visual Control Reacher Hardreturn: 53.8 -> 466.0Improved
apply TaskLoom instead of single-task DreamerV3DreamerV3 · Proprio Controlmean return: 456.1 -> 647.4Improved
apply TaskLoom instead of single-task DreamerV3DreamerV3 · Visual Controlmean return: 382.0 -> 455.3Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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