花見🌸

RT @PalettaAnthony: Nakagin exhibit at MoMA both great and depressing

It wasn't even the result of a bad prompt, it's in the training 😭.

RT @Hesamation: a guy created a dataset of 50 books from London 1800-1850 for LLM training. no modern bias. it’s actually super cool to se….

It's kinda wild to see Claude intentionally send erroneous calls to an API in order to receive the correct request format because it couldn't find enough information online to craft the correct response initially. .

Kioi Seido.紀尾井清堂

I really would like to know *why* you would degrade yourself by working at xAI? This is possibly the most impactful time to be working in AI and you're working on. this?.

RT @jeremyphoward: I replicated this result, that Grok focuses nearly entirely on finding out what Elon thinks in order to align with that,….

RT @uhe_okmt:

RT @NighSide: Elon Musk: We have improved Grok yay!. Grok: I am now MechaHitler & MechaHitler endures

RT @fchollet: In order to supervise an automation tool (or another person!) effectively, you need to be able to do the same job yourself.….

蝉がようやく目を覚ました.

RT @ukii92: snap

RT @rasbt: If you're getting into LLMs, PyTorch is essential. And lot of folks asked for beginner-friendly material, so I put this together….

わーーーーーーい.

Ban the mild air-conditioned carriages in the summer. Nobody wants mild aircon during rush hour.

RT @gabriberton: Here's a cool paper using LLMs for lossless text compression, in what they call LLMZip, which outperforms SOTA text compre….

RT @chowtingagnes: 無事に大学院を卒業しました。. 宿題が多くて疲れた時もあったけど、優しい先生やクラスメイトに出会えて、色々学ぶことができて楽しかったです。実感がまだないですが、一応修士卒業生です！

RT @drmichaellevin: @TedPavlic yes!! Here I argue that creative interpretation of memories (very close to confabul….