Alexander Grishin

[Alexander Grishin @AlexGrishin_]

@SchoeneggerPhil @SpencrGreenberg @LuciusCaviola Wrong Alexander mate

[Alexander Grishin @AlexGrishin_]

@Nekrolm Everyman очень приятный

[Alexander Grishin @AlexGrishin_]

@TmlrOrg Thank you for your great initiative!
Could you please clarify, does TMLR have or inherit from JMLR the rank? Or not yet?

[Alexander Grishin @AlexGrishin_]

@norpadon Он не лабу возглавлял в airi, а скорее консультировал по текущим/планируемым проектам изредка, но с началом войны открестился, конечно

[Evgenii Nikishin @nikishin_evg]

Excited to share the DARL workshop @darl_icml at the upcoming ICML 2022 @icmlconf!

Check out the great line of speakers and please consider contributing your RL papers!

[Nikita Lisitsa @lisyarus]

@norpadon Yep, adding second-order terms was my first impulse. Looks like I underestimated the number and size of required layers :)

[Alexander Grishin @AlexGrishin_]

@norpadon В деепричастные?

[AIRI (Artificial Intelligence Research Institute) @AIRI_inst]

Congratulations to the team of our RL group team-lead @d_l_l_l_b on winning the Nethack Challenge @aicrowdHQ at the world's oldest machine learning conference NeurIPS!
https://t.co/NpIwyyI1gL

[Alexander Grishin @AlexGrishin_]

@norpadon 4'ый будешь пробовать?

[Alexander Grishin @AlexGrishin_]

@makezur

[Alexander Grishin @AlexGrishin_]

@AllDeepLearning @ak92501 @_arohan_ It's actually a more or less usual way of running for humanoid-like RL agents

[Alexander Grishin @AlexGrishin_]

@vasha_zanoza Есть ещё такая интересная история про посадку
https://t.co/tYee5wy1PG

[Alexander Grishin @AlexGrishin_]

@norpadon Мой друг у Бурнаева коллаборирует с вроде крутым геометром (https://t.co/ivb769YDc7), но я не сварщик

[Alexander Grishin @AlexGrishin_]

Meta-learning on MNIST/Omniglot turns out to be really fun
0, 1, ..., 9

[Alexander Grishin @AlexGrishin_]

@norpadon 9 болезней грядок из 10

[arxiv @arxiv_org]

Controlling Overestimation Bias with Truncated Mixture of Continuous Distributional Quant... https://t.co/av7I9ctqbQ

[Pavel Shvechikov @shvechikov_p]

Our new paper, Truncated Quantile Critics, improves SOTA on MuJoCo by 20-30% ! With TF and PT code.

Credit to @brickerino @shvechikov_p @AlexGrishin_ and Dmitry Vetrov.

Video: https://t.co/o0zvcuHa8X
Project page: https://t.co/ppZi3Tc1VJ
Paper: https://t.co/XodndN2enO

[Alexander Grishin @AlexGrishin_]

Workshop "#RL under Partial Observability", Room 517 C
"Joint Belief Tracking and Reward Optimization through Approximate Inference"

Come and discuss with me our work!
#NeurIPS2018 #bayesgroup

[Bayesian Methods Research Group @bayesgroup]

Workshop "#RL under Partial Observability", Room 517 C
Contributed talk at 10:10 AM, just after invited talk from David Silver.

Paper: https://t.co/vSI73XPbWW
By Pavel @shvechikov_p Shvechikov, Alexander @AlexGrishin_ Grishin, Arseny Kuznetsov, Alexander Fritzler, Dmitry Vetrov

[Alexander Grishin @AlexGrishin_]

“Jeff Dean compiles and runs his code before submitting, but only to check for compiler and CPU bugs” - good old Chuck pales in comparison
https://t.co/FJqbbxdq7b