The name is the button.

An agent writes the code now, and a human thumb hovers over Accept all a hundred times a day. The honest question of this decade is what, exactly, you just accepted. So I read the diff.

Eight years as a backend engineer in Stockholm, back in Tokyo since 2023, and I still ship something small most nights, usually badly. My one rule: the only real review of a model is the thing you built with it. That is why there is no leaderboard on this site.

Around a hundred and forty sources come through — the labs' own posts, the papers, the tools, the runtimes and languages nobody is watching yet, the chip under the demo, and the newsletters people actually forward. What is moving gets counted so the week reads at a glance, and every few days I explain one mechanism in plain words.

A story here is a headline, a short excerpt, one image and the link to whoever wrote it. We never reproduce the post; the images belong to their authors. A post written in Japanese appears in Japanese, exactly as it was published, and the link goes to the original.

— Yuki Halvorsen, Tokyo

← Accept All Archive

llama.cpp releasesTools

b11453: llama: remove the gather path of the glm5-next sparse attention (#30042)

October 6

The gather path attended over the selected latents with a plain matmul and softmax. It only ran with n_ubatch <= 16, and the flash attention backends now skip the masked rows through n_kv_max, so…

Models & releases

Read at llama.cpp releases ↗

One letter a week, in your inbox.

Latest

b11453: llama: remove the gather path of the glm5-next sparse attention (#30042)

Related

GitHubEmbeddingGemma 2: An open, lightweight multimodal embedding model Hacker News (front)

ValleyAnthropic is giving startups a free year of Claude Team and $1,000 in credits TechCrunch AI

EuropeDelivery Hero alumni have built 150 startups worth €7.5B, making it Europe’s biggest founder factory Tech.eu

VoicesScrimshaw Jukebox Simon Willison

ValleyMistral’s new 1T model aims to leapfrog closed and open rivals TechCrunch AI

EuropeMistral unveils le Chonk, says "significantly outperforms” any open-weight models developed in US or Europe Tech.eu