msuzen.bsky.social - Profile | ThreadSky | a Reddit-style client for Bluesky

#ReinforcementLearning #RL Congrats to Barto & Sutton. Actually this. one is also physics domain. I wouldn't surprise if they are nominated. for Physics Nobel too. awards.acm.org/about/2024-t...

submitted 15 days ago • 0 comments

Bekenstein-Landauer ultimate data storage limit science-memo.blogspot.com/2025/04/ulti... #Cosmology #InformationTheory #StatisticalPhysics

submitted 32 days ago • 0 comments

Yes. Zero is even #maths

submitted 34 days ago • 0 comments

Pitfalls of exponential function: Statistical Physics to Machine Learning memosisland.blogspot.com/2025/04/soft... #MachineLearning #StatisticalPhysics cc @stephenwolfram.bsky.social

submitted 38 days ago • 0 comments

Entia non sunt multiplicanda praeter necessitatem Occam #Overfitting #DataScience #MachineLearning

submitted 51 days ago • 0 comments

The concept of Occam's gap Resolution of misconception of overfitting: Differentiating learning curves from Occam curves science-memo.blogspot.com/2023/04/Occa... #Overfitting #MachineLearning #DeepLearning #AI

submitted 53 days ago • 0 comments

Using JWST, researchers have monitored an isolated exoplanet for 2.5 of its rotation periods. The object’s variable IR spectrum confirmed previous hints of patchy clouds and provided new evidence of hot spots in the upper atmosphere and fluctuating carbon chemistry.

submitted 59 days ago • 0 comments

Very excited about the upcoming Natural Philosophy Symposium in Baltimore, May 29-31. There still some room to register! www.naturalphilosophyhopkins.org/natural-phil...

submitted 63 days ago • 4 comments

"A mathematician is a machine for turning coffee into theorems" Paul Erdös #LongCoffee #Maths #AppliedMaths #physics

submitted 65 days ago • 0 comments

In-Context Learning meets Chaos In-context learning is quite promising but limitations appear akin to chaotic dynamics. A new Kind of AutoML: memosisland.blogspot.com/2024/05/llm-... #llm #reasoning #InContextLearning #chaos #ai #ChainOfThought #LLMstability #SymbolicAutoML

submitted 68 days ago • 0 comments

Brilliant article on Game of Life from Legendary Dr. Wolfram . So rich both mathematically and from statistical physics perspective. cc @stevenstrogatz.com

submitted 73 days ago • 0 comments

Focus session on Quantum Machine Learning @apsphysics.bsky.social summit.aps.org/events/MAR-X...

submitted 74 days ago • 0 comments

Now _that's_ a dish towel.

submitted 82 days ago • 2 comments

Games and statistical physics Entropy production is bound to have a critical temperature that after it doesn’t change. This insight is shown for the first time with such a simple system without resorting to free-energy computations. #statisticalphysics #entropy #games arxiv.org/abs/2503.03769

submitted 83 days ago • 1 comment

Start you causal machine learning journey with the looper repo today: Looper : A resource list for causality in statistics, data science and physics github.com/msuzen/looper

submitted 128 days ago • 0 comments

Lyapunov-Conjecture for CoT #llms #ai In-context learning as a new kind of symbolic-AutoML: Lyapunov conjecture for CoTs. Available from: www.researchgate.net/publication/...

submitted 161 days ago • 0 comments

This type of logical Quantum Error Correction (QEC) is promising indeed #quantum #computing

submitted 162 days ago • 0 comments

(1) Repeatability: Same team, same experimental setup (2) Reproducibility: Different team, same experimental setup (3) Replicability: Different team, different experimental setup [*] www.acm.org/publications... cc @odedrechavi.bsky.social #datascience #ai #reproducibility

submitted 171 days ago • 0 comments

Nonlinear Dynamics and Chaos A must have course for quantitative scientists and practitioners, critical thinking skills. #ai #datascience #appliedMaths #physics www.youtube.com/playlist?lis... cc @stevenstrogatz.com @sfiscience.bsky.social

submitted 171 days ago • 1 comment

We don’t want to stifle innovation, but we do need checks and balances, if we are to get to a positive AI future. Gary Marcus @garymarcus.bsky.social Taming Silicon Valley How We Can Ensure That AI Works for Us, (MIT Press 2024) mitpress.mit.edu/978026255106... #ai #ml

submitted 173 days ago • 0 comments

hot take: scientists are great people and should help each other live the best life possible

submitted 176 days ago • 3 comments

Metric tensor demystified cc @stevenstrogatz.com @stephenwolfram.bsky.social science-memo.blogspot.com/2024/04/metr...

submitted 184 days ago • 0 comments

Reversibility paradox meets causality with cellular automata cc @stephenwolfram.bsky.social H-theorem do-conjecture: arxiv.org/abs/2310.01458 Ising-Conway Games: memosisland.blogspot.com/2023/10/isin... #causality #gameOfLife #reversibilityParadox

submitted 192 days ago • 0 comments

Understanding combinatorics via cartesian products It is amazing that repeated cartesian products are sufficient to count in complex situations cc @stevenstrogatz.bsky.social memosisland.blogspot.com/2024/09/comb...

submitted 192 days ago • 0 comments

Causality: A new science of data looper : A resource list for causality in statistics, data science and physics github.com/msuzen/loope... #causality #causalAI #inference

submitted 192 days ago • 0 comments