Mehtaab Sawhney (@mehtaab_sawhney)

Vishesh Jain과 Clayton Mizgerd가 GPT-5.4 Pro를 사용해 마코프 체인 이론의 중요한 결과를 증명한 최근 논문이 소개되었습니다. 최신 AI 모델이 수학/이론 연구에 활용된 사례로, 고난도 정리 증명에 생성형 모델이 쓰였다는 점이 주목됩니다.

https://x.com/mehtaab_sawhney/status/2041354267286737243

#gpt54pro #markovchains #mathresearch #llm #ai

Mehtaab Sawhney (@mehtaab_sawhney) on X

In a beautiful recent paper, Vishesh Jain and Clayton Mizgerd used GPT-5.4 Pro to prove a striking result in the theory of Markov chains: https://t.co/KU8nGjHiB7 They study the adjacent transposition Markov chain on the symmetric group. A conjecture of Fill, recently settled by

X (formerly Twitter)

Where the fun starts was to cover the markov chains aspect. Keeping this ultra simple, the decision will work as follow. I flipped a coin 48 times and wrote the results. That will be my "score".

I am starting at middle C and, at each bar, for Head, I'll go up, and for Tails, I'll go down. I'll only use the C Major scale. Once written down, I played this in real time on my Ensoniq VFX SD.

#markovchains #music #math #probabilities #electronica #live #jamuary #jamuary2026

Patternuary 13 (2024)

PeerTube

Beginner #MarkovChains question:
I'm working on a small analysis program for a friend. It's already existing code where you select a number of labels from a labelled sequence of interactions and then supposedly get a transition matrix from them.

BUT - the code ignores intermediate labels when counting transitions.
e.g. if a sequence has three labelled members w,x,y, and you only want to analyse w, and y, the current code would count that as a transition from w to y, ignoring the intervening x.

This does not seem appropriate for calculating transition probabilities. If I understand correctly, this should not be counted at all, since the transition was w -x, then x-y.

Does that sound like I have understood correctly?

🚨 BREAKING: High school essay from 2012 declares *Markov Chains* as the OG language models! 🙄 Apparently, *Linear Algebra 101* had all the answers before #AI got cool. Who knew teenage musings could redefine the hype cycle? 😂
https://elijahpotter.dev/articles/markov_chains_are_the_original_language_models #MarkovChains #LinearAlgebra #HypeCycle #TeenageMusings #HackerNews #ngated
Markov Chains Are the Original Language Models

Back in my day, we used math for autocomplete.

Markov Chains Are the Original Language Models

Back in my day, we used math for autocomplete.

What I’ve been reading (and watching) this week ending 24 August 2025

Lessons from U.S. Army Special Ops on Becoming a Leader

Medium
The Strange Math That Predicts (Almost) Anything

YouTube
The Strange Math That Predicts (Almost) Anything

YouTube
2025-08-11 08-26-10 - Monday Morning Practice

PeerTube