sacha2.bsky.social
@sacha2.bsky.social
Pinned
I might look smart, however, I am absolutely not.
Reposted
Salam Afiouni, Jakub Cerny, Chun Kai Ling, Christian Kroer
Colonel Blotto with Battlefield Games
https://arxiv.org/abs/2511.06518
November 11, 2025 at 6:28 AM
1h 40 min of Danijar Hafner open.spotify.com/episode/6Qbb...

the guy who almost made me cry multiple times
Danijar Hafner on Dreamer v4
open.spotify.com
November 10, 2025 at 8:16 PM
so many interesting (for me, personally) papers at this neurips
November 9, 2025 at 8:37 PM
1 thought away from emailing gabrielle farina with a question
November 9, 2025 at 8:34 PM
Reposted
- Other minds by @petergs.bsky.social
- Ways of being by @jamesbridle.bsky.social
- Immune by Philipp Dettmer
- The self-assembling brain by Robin Hiesinger
- What is intelligence by @blaiseaguera.bsky.social
- Life as no one knows it by @saraimari.bsky.social
- Thread ripper by Amalie Smith
November 9, 2025 at 5:01 PM
I'd put myself in that wacky position when I became interested in a very narrow list of topics. I really found not like a single person to talk about and thus, develop into the topic except of reading some papers.

it's like trapping myself in the box
November 7, 2025 at 8:25 PM
Apply!
I'm hiring a student researcher for next summer at the intersection of MARL x LLM. If you're a phd student with experience in MARL algorithm research, please apply and drop me an email so that I know you've applied! www.google.com/about/career...
Student Researcher, PhD, Winter/Summer 2026 — Google Careers
www.google.com
November 7, 2025 at 7:07 AM
@natashajaques.bsky.social , may I participate in a project; not as a student, I mean, but publicly, on this platform, for your subjective judgement? 😇
found a fresh vid for you www.youtube.com/watch?v=Yz58...

I think you can ask @natashajaques.bsky.social some questions, she seems to be knowledgeable of the said subject 💅
November 6, 2025 at 9:16 PM
i do not want to work from home

it's tough
November 6, 2025 at 7:27 PM
may I say that I enjoy analogies that came from neuro-, cognitive or behavioural sciences, like Q-learning, Bayesian ToM, but not the sciences themselves. I am very critical of them.

If you knew... how it hurts me to explain RL and model-based methods from that pov. Maybe for good.
November 6, 2025 at 5:10 PM
sounds like kind of good news, doesn't it, @sharky6000.bsky.social ?
November 6, 2025 at 4:30 PM
Reposted
Introducing Petri Dish Neural Cellular Automata (PD-NCA)

pub.sakana.ai/pdnca/

In this work we explore the role of continual adaptation in artificial life, where the cellular automata in our system do not rely on a fixed set of parameters, but rather learn continuously during the simulation itself.
November 5, 2025 at 12:26 AM
people bank on self-play, until they hear of its convergence properties
November 4, 2025 at 7:38 PM
Reposted
New #J2C Certification:

Synthesizing world models for bilevel planning

Zergham Ahmed, Joshua B. Tenenbaum, Chris Bates, Samuel J. Gershman

https://openreview.net/forum?id=m9V4JHLJrD

#planning #reinforcement #games
November 3, 2025 at 9:24 PM
Reposted
I would be interested in:
1. Putting those ideas to test and measure the reduction in the Alignment Gap

2. Scale the complexity of the environment (e.g. Age of Empires but without hardcoded property rights) and scale the capabilities of the Agents (e.g. use super tiny LLMs).
November 3, 2025 at 5:44 PM
. @sharky6000.bsky.social have you ever played this game?
I certainly did, and completely forgot about it. Looks like a simple yet challening environment
November 3, 2025 at 5:49 PM
BRO
November 1, 2025 at 7:06 PM
have you seen this cheeky open-spiel wrapper, @sharky6000.bsky.social

colab.research.google.com/github/meta-...

it's a quite interesting concept of client-server architecture
Google Colab
colab.research.google.com
November 1, 2025 at 6:37 PM
Reposted
Check out this blog post on Kinship-aligned MARL!

abranti.com/the-key-prop...

Shared by author Joao Abrantes.

See Part 2 below 👇
Emerging a Society with MARL Part 1: The Key Properties to Reproduce Society with Multi-Agent RL
We introduce an Environment that benefits agents who can cooperate — despite having different goals. We argue that these were the conditions that shaped human social capabilities.
abranti.com
October 29, 2025 at 7:52 PM
rofl
October 29, 2025 at 6:02 PM
Battleship arena for LLMs

another cool imperfect information task

www.gabegrand.com/battleship/

cc @sharky6000.bsky.social
BattleshipQA: Shoot First, Ask Questions Later?
BattleshipQA
www.gabegrand.com
October 28, 2025 at 9:09 PM
bluesky is for american type of people

uncomfortable
October 28, 2025 at 8:53 PM