CoLLAs 2026: Michael Bowling challenges MDP foundations of continual reinforcement learning

apsarathchandar · x · 2026-09-15

In the first CoLLAs 2026 Day 1 keynote, Michael Bowling challenged the core foundations of Continual RL: are MDPs, optimal policies, and ergodicity assumptions holding the field back? The talk explored alternative foundations and new algorithms for ever-changing worlds.

Original post →

More from Research

Research channel →