Maximilian Schlegel
@schlegelm.bsky.social
Research Scientist @ Google, working on RL & sequence models
Reposted by Maximilian Schlegel
Well this is exciting: arxiv.org/abs/2512.20605
Emergent temporal abstractions in autoregressive models enable hierarchical reinforcement learning
Large-scale autoregressive models pretrained on next-token prediction and finetuned with reinforcement learning (RL) have achieved unprecedented success on many problem domains. During RL, these model...
arxiv.org
January 6, 2026 at 7:53 PM
Well this is exciting: arxiv.org/abs/2512.20605
Check out João Sacramento’s talk about our work on “Internal Reinforcement Learning”!
Here’s the link to the recording: youtu.be/cx_MIhvAOYM?...
And here’s the link to the new paper (arxiv.org/abs/2512.20605)
Here’s the link to the recording: youtu.be/cx_MIhvAOYM?...
And here’s the link to the new paper (arxiv.org/abs/2512.20605)
Joao Sacramento - “Emergent temporal abstractions in autoregressive models enable hierarchical RL”
YouTube video by MAIN Conference
youtu.be
December 26, 2025 at 6:12 PM
Check out João Sacramento’s talk about our work on “Internal Reinforcement Learning”!
Here’s the link to the recording: youtu.be/cx_MIhvAOYM?...
And here’s the link to the new paper (arxiv.org/abs/2512.20605)
Here’s the link to the recording: youtu.be/cx_MIhvAOYM?...
And here’s the link to the new paper (arxiv.org/abs/2512.20605)
Reposted by Maximilian Schlegel
Seijin Kobayashi, et al.: Emergent temporal abstractions in autoregressive models enable hierarchical reinforcement learning https://arxiv.org/abs/2512.20605 https://arxiv.org/pdf/2512.20605 https://arxiv.org/html/2512.20605
December 24, 2025 at 6:34 AM
Seijin Kobayashi, et al.: Emergent temporal abstractions in autoregressive models enable hierarchical reinforcement learning https://arxiv.org/abs/2512.20605 https://arxiv.org/pdf/2512.20605 https://arxiv.org/html/2512.20605
Super cool paper
Weight decay induces low-rank attention layers
The effect of regularizers such as weight decay when training deep neural networks is not well understood. We study the influence of weight decay as well as $L2$-regularization when training neural ne...
arxiv.org
December 11, 2024 at 2:12 PM
Super cool paper
Reposted by Maximilian Schlegel
Reconnecting with the Twitter homies on Bluesky
November 20, 2024 at 7:32 AM
Reconnecting with the Twitter homies on Bluesky
detailed results of this are available at slatestarcodex.com/Stuff/aituri.... quickly checked for the top5 pictures among those who hate ai art. The two winning ones are AI (3,4,5: human) but close in style to famous impressionist paintings as others pointed out
November 21, 2024 at 9:12 AM
detailed results of this are available at slatestarcodex.com/Stuff/aituri.... quickly checked for the top5 pictures among those who hate ai art. The two winning ones are AI (3,4,5: human) but close in style to famous impressionist paintings as others pointed out
let’s goo! gpu cluster secured
You made the nice list Max. If you’re going to go, go big.
Happy holidays 🎅🏻
Happy holidays 🎅🏻
November 20, 2024 at 8:59 PM
let’s goo! gpu cluster secured
Dear Santa,
My name is Max, I’m 23 years old and live in Zurich. For Christmas I would love to find a
24xH100 GPU cluster
under the tree!
Sincerely,
Max
My name is Max, I’m 23 years old and live in Zurich. For Christmas I would love to find a
24xH100 GPU cluster
under the tree!
Sincerely,
Max
November 20, 2024 at 7:19 PM
Dear Santa,
My name is Max, I’m 23 years old and live in Zurich. For Christmas I would love to find a
24xH100 GPU cluster
under the tree!
Sincerely,
Max
My name is Max, I’m 23 years old and live in Zurich. For Christmas I would love to find a
24xH100 GPU cluster
under the tree!
Sincerely,
Max