← Search

Mert Albaba

2 accepted papers

2026

Test-time Offline Reinforcement Learning on Goal-related Experience

ICML 2026poster

Foundation models compress a large amount of information in a single, large neural network, which can then be queried for individual tasks. There are strong parallels between this widespread framework and offline goal-conditioned reinforcement learning algorithms: a universal value function is train…

Cited by 0SourceScholar