Lots more here: This was a ton of work from an awesome team. Thanks to @yuqi_xie5, @JustinSasek, Steven Zheng, and @yukez.

Competitive Pokémon has all the makings of a great RL research problem:. - Stochasticity? ✅.- Imperfect information? ✅.- Generalization? ✅.- More fun than half-cheetah? ✅✅✅. And today it gets better…

RT @yukez: We took a short break from robotics to build a human-level agent to play Competitive Pokémon. Partially observed. Stochastic. Lo….

There’s an RL trick where we turn Q-learning into classification. Among other things, it’s a quick fix for multi-task RL’s most unnecessary problem: that the scale of each task’s training loss evolves unevenly over time. It’d be strange to let that happen in supervised learning,

RT @yukez: One of RL's most future-proof ideas is that adaptation is just a memory problem in disguise. Simple in theory, scaling is hard!….

RT @arXiv_Daily: Measuring Visual Generalization in Continuous Control from Pixels.by Jake Grigsby et al. #Supervis….

RT @jxmorris12: An updated version of the #TextAttack paper has posted, including:. • Implementations of 16 papers on NLP attacks.• Dozens….