shubhamcore – Telegram

shubhamcore

5 subscribers

51 photos

1 file

85 links

Download Telegram

About

Blog

Apps

Platform

helppp i said deprecation instead of depreciation sksksksskjjjs

7 views20:04

freudian slip or sm 😭

7 views20:05

lmao. opened this after years

❤1

7 views20:09

I believe P=NP
https://emanueleviola.wordpress.com/2018/02/16/i-believe-pnp/

i wish i understood the examples better lol

The only things that matter in a theoretical study are those that you can prove, but it’s always fun to speculate. After worrying about P vs. NP for half my life, and having carefully reviewed the …

5 views05:19

Forwarded from tomrum

Monotony, my constant woe, ally and foe

4 views02:49

https://qwen-neurons-site.vercel.app/#qwen/c-L9-9477-m-L21-1037-m

qwen-neurons-site.vercel.app

Neuron Persona Library

Each of Qwen3-8B’s 442,368 MLP neurons, pushed alone up and down, then asked “Who are you?”. Search the replies or open any neuron.

4 views12:00

me

4 views12:34

5 views15:43

4 views08:19

https://www.youtube.com/watch?v=ZRfCzEAOdUQ
INSANE

A Breathtaking 17th-Century Hacienda Filled With Art and Antiques | House Tour

To SHOP items inspired by our home tours, visit https://shopmy.us/shop/homeworthy

Artists Laura Kirar and Richard Frazier take us inside Hacienda Subin, their extraordinary 17th-century home in Mérida, Yucatán, which they discovered as a jungle-covered ruin…

4 views09:28

4 views12:51

https://blog.cosmos-institute.org/p/you-cant-speedrun-yourself
"Cheap intelligence poses a challenge to our longstanding model of formation, in which judgment is acquired through perseverance and apprenticeship. AI can increasingly offer people an approximation of the fruits of determined work, without them having to endure the hard work."

blog.cosmos-institute.org

You Can’t Speedrun Yourself

The myth of high-level work

4 views13:01

Forwarded from Judah

https://x.com/IvanVendrov/status/2115720125328634247?a=MF0m234S2XAUF02Lbmw&s=20

X (formerly Twitter)

ivan (@IvanVendrov) on X

Good reminder that ~all of generative AI is in some sense adversarially optimized for you to think it is better than it is (via pre-training, via RLHF, via hill-climbing benchmarks). or as nostalg…

4 views14:03

omg
https://rikiyat.github.io/blog/weight-space-programming/

rikiyat.github.io

Sparks of Weight-Space Programming: LLMs can natively read & write neural network weights

We find that frontier LLMs are surprisingly good at natively reading and writing neural network weights, using a variety of recent tools and algorithms or even via direct weight editing.

4 views04:14

love this one
https://www.experimental-history.com/p/thank-you-for-being-annoying

Experimental-History

Thank you for being annoying

OR: whack 'em if you got 'em

4 viewsedited 08:00

he jus like me fr

🌚1

5 views12:54

https://sshhole.com/

*sshole · sshhole.com

openssh on musl linux has been letting people into password-less accounts since 2016.

4 views17:50

any good blazing fast mac and android notes app?

3 views10:25

A proof of the Aaronson-Ambainis conjecture
https://psi.inc/research/aaronson-ambainis/

A proof of the Aaronson-Ambainis conjecture

A proof of the Aaronson-Ambainis conjecture about quantum speedups on random inputs, with a complete formalization in Lean.

3 views17:53

anyone can help me see when multiplexers are needed? really not seeing the point

3 views19:19