AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

Cross-Episodic Curriculum for Transformer Agents

Lucy Xiaoyang Shi, Yunfan Jiang, Jake Grigsby, Linxi Fan, Yuke Zhu

NeurIPS 2023 · Reinforcement learning › Visual and representation RL · approach: Curriculum and environment design · 1 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 cross-episodic attention (ablation)CEC Transformer agent · DMLab Goal Mazesuccess rate: 35.0 -> 65.2Improved
add cross-episodic attention (ablation)CEC Transformer agent · DMLab Watermazesuccess rate: 20.0 -> 50.9Improved
add cross-episodic attention (ablation)CEC Transformer agent · DMLab Irreversible Pathsuccess rate: 3.8 -> 38.2Improved
add cross-episodic attention (ablation)CEC Transformer agent · RoboMimic Liftsuccess rate: 75.9 -> 100.0Improved
add cross-episodic attention (ablation)CEC Transformer agent · RoboMimic Cansuccess rate: 99.3 -> 100.0No clear effect
use fine-grained instead of coarser task-difficulty curriculum (ablation)CEC (Task Difficulty) · DMLab levelsrelative performanceImproved
use a moderate context length instead of too short or too long (ablation)CEC (Task Difficulty), Auto · DMLab levelsrelative performanceMixed
cross-episodic curriculum with task-difficulty vs single-difficulty ATCEC (Task Difficulty), Auto · DMLab (avg of 3 levels)average score: 42.7 -> 51.4Improved
cross-episodic curriculum vs DT with mixed difficulty dataCEC (Task Difficulty), Auto · DMLab (avg of 3 levels)average score: 35.3 -> 51.4Improved
fixed task sequencing instead of automatic promotionCEC (Task Difficulty) · DMLab (avg of 3 levels)average score: 51.4 -> 54.4Improved
learning-progress curriculum instead of task-difficulty curriculumCEC · DMLab (avg of 3 levels)average score: 51.4 -> 32.4Hurt
CEC vs BC on expert data onlyCEC (Task Difficulty), Auto · DMLab (avg of 3 levels)average score: 14.2 -> 51.4Improved

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