Jan

die amerikanische bandenkultur ist so dumm

ihr wollt mich umbringen, weil ich einen tanz gemacht habe, den ich nicht machen sollte? ich darf diese straße nicht entlanggehen, weil ich blau trage? Get a job

Do given the tables
https://gist.github.com/bytemouse/51615a6ddb82d7fc7aac35f65cb98902

The loop takes like 3 to 5 cycles per iteration depending on branch prediction

How do I actually know how long each operation in assembly takes roughly? I have no idea but would like to know
https://godbolt.org/z/eK7cdjqhs

Belated Valentine's Day post

Few such cases

Eine Menge Leute auf dieser Plattform die wollen das ich andere Menschen Dinge genießen lasse. Bitte versteht, dass ich niemanden etwas genießen lassen will und werde smh

Eine schwarz-weiße Katze liegt mit einem Plüschmaus auf einem gemusterten Sessel.

Ich und wer?

meinen Schwanz nicht lutschst, hinterlässt es Narben in meiner Seele

The image is from a "Transparency Center" document and lists guidelines regarding acceptable and prohibited content for insults. It mentions: 1. Insults about: Character, such as cowardice, dishonesty, criminality, and sexual promiscuity or immorality. Mental characteristics, including but not limited to accusations of stupidity, intellectual capacity, and mental illness, as well as unsupported comparisons among politically correct (PC) groups based on inherent intellectual traits. 2. Highlighted section: The document allows allegations of mental illness or abnormality when tied to gender or sexual orientation, referencing political and religious discourse about transgenderism and homosexuality. It also acknowledges the non-serious use of terms like "weird."

Meta literally created a LGBTQ exception for calling someone mentally ill as an insult. You can't do it for any other group except LGBTQ people.

Just three social media CEOs posting on their platform today.

I feel like it definitely looks more like a story

What is a picture that you actually like?

If you're worried about the New Year, remember that this little Teddiursa believes in you! 🐻

I was just thinking that to create a Lean-optimized fine-tuned LLM, I would just need to hire an army of students to create a labeled dataset of half a million Lean proofs and secure one million dollars to spend on gpu hours.

https://daringfireball.net/2024/12/openai_unimaginable

>OpenAI’s Board, Paraphrased: ‘To Succeed, All We Need Is Unimaginable Sums of Money’

Same tbh

if you really dedicate yourself to an instrument you can, over many years of rigorous practice, become so skilled that you can only produce music that is completely unlistenable

Schizo geo guessing posting in the eyes of the uninitiated

What the fuck are you even talking about

adam.math.hhu.de

Try Lean by creating the natural numbers with the peano axioms, it's pretty fun.

A friend talked to me about how about LLMs can't solve tasks for the mathematical logic lecture but given an appropriate set of axioms and integration with lean most current LLMs easily can solve all math exercises in university.

www.ams.org/notices/2025...

I'd call this paper a "big deal" in that it is a normalization of, very fair summary of, and indication that there is a future for, LLMs in pure mathematics from one of its leading practitioners.

https://adam.math.hhu.de/

Try Lean by creating the natural numbers with the peano axioms, it's pretty fun.

solve tasks for the mathematical logic lecture but given an appropriate set of axioms and integration with lean most current LLMs easily can solve all math exercises in university.

Incel lingo is legitimately hilarious

If you try to get girls by being funny and entertaining it's called jestermaxxing.

like a much less impressive step-function increase.

An important caveat of the ARC-AGI results is that the version of o3 they evaluated was actually trained on a public ARC-AGI training set. By contrast, to my knowledge, none of the o1 variants (nor Claude) were trained on said dataset.

https://arcprize.org/blog/oai-o3-pub-breakthrough

This is a line graph showing AI model performance scores in arc-agi from 2019-2025. Starting with GPT-2 at near 0%, there's a modest increase through GPT-4, followed by a dramatic upward curve reaching ~90% with unreleased "o3 tuned high" models. At the end of the graph is a brick wall humorously referencing the sentiment that AI hit a wall.

LLM Progress hit a wall

A Timex Expedition watch with leather strap shown on a wrist, photographed against a car's steering wheel which has a yellow circular sticker featuring a wonky Ferrari logo.

At this time of year, it is important to remember what we have and be grateful

This is a meme consisting of three images arranged horizontally: 1. A cartoon/anime-style illustration of a person with long black wavy hair wearing a pink top with a Hello Kitty design and looking somewhat distressed 2. A middle image showing an overflowing laundry basket full of unfolded clothes 3. An illustration depicting Jesus wearing a crown of thorns and a halo The text overlaid on the images reads: "pls stop giving me your toughest battles" (on the left) "you literally just need to fold your clothes" (on the right)

Literally me