publications

2026

1. ICLR 2026

   

   TRACED: Transition-aware Regret Approximation with Co-learnability for Environment Design

   Keywords #Curriculum Learning, #Reinforcement Learning, #Unsupervised Environment Design, #Regret Approximation, #Meta-Learning, #Generalization

   Geonwoo Cho, Jaegyun Im, Jihwan Lee, Hojun Yi, Sejin Kim, and Sundong Kim

   In ICLR , 2026

   Also presented at CoRL-25 Rational Robots Workshop

   arXiv Code Website
2. ICMLw2026

   Receding-Horizon Execution for Action Chunking in Offline-to-Online Reinforcement Learning

   Jihwan Lee, Geonwoo Cho, Hojun Yi, Jaegyun Im, Giseung Park, and Sundong Kim

   In ICML Workshop on Decision-Making from Offline Datasets to Online Adaptation: Black-Box Optimization to Reinforcement Learning , 2026