1681
Executing programs inside transformers with exponentially faster inference
We build a computer inside a transformer — executing arbitrary C programs for millions of steps with exponentially faster inference via 2D attention heads.
Top posts
Every blog post that took off on Hacker News, ranked by the votes of its best submission. 6,326 posts.
1682
Running Gemma 4 26B at 5 tokens/sec on a 13-year-old Xeon with no GPU
A storage appliance that shipped before AVX2 existed now runs Google's Gemma 4 26B in my basement, no GPU. Here's how we got a modern MoE model onto a CPU older than the instructions its inference engine assumed were there.
1683
1684
1685
The Algorithms Behind Probabilistic Programming
We recently introduced our report on probabilistic programming. The accompanying prototype allows you to explore the past and future of the New York residential real estate market. This post gives a feel for the content in our report by introducing the algorithms and technology that make probabilistic programming…
1686
Bag of words, have mercy on us
OR: Claude will you go to prom with me?
1687
Early progress on accelerating drug development with Claude
Over the past several months, we’ve partnered with six organizations to accelerate curing disease. We’ve also started drug discovery efforts of our own, including programs for rare childhood forms of muscular dystrophy. In this post, we provide an update on our progress.
1688
Grok 4
Released last night, Grok 4 is now available via both API and a paid subscription for end-users. Update: If you ask it about controversial topics it will sometimes search X …
1689
OpenAI is well positioned to fast-follow Jev
TypeSafe's Jev is a genuine breakthrough – snap judgments with calibrated probabilities instead of generated text. My bet is OpenAI is already figuring out how to copy it, and then embed it inside its own models where Jev can't follow.
1690
Why DeepSeek is cheap at scale but expensive to run locally
Why is DeepSeek-V3 supposedly fast and cheap to serve at scale, but too slow and expensive to run locally? Why are some AI models slow to respond but fast once…
1691
Weird A.I. Yankovic: a cursed deep dive into the world of voice cloning
What would it sound like if the musicians parodied by "Weird Al" Yankovic sang *his* parodies instead? I jumped off the deep end into the world of A.I. cover songs to find out, and the results are even worse than you imagine.
1692
1693
So you want to build your own open source chatbot
A small team within Mozilla’s innovation group recently undertook a hackathon to build a trustworthy internal chatbot prototype.
1694
NotebookLM Audio Overviews are now available in over 50 languages
Learn more about NotebookLM’s Audio Overviews feature and its expansion to more than 50 languages.
1695
1696
Don't bother parsing: Just use images for RAG
If search is the game, looks matter
1697
Attacks against machine learning – an overview
This blog post surveys the attacks techniques that target AI (Artificial Intelligence) systems and how to protect against them.
1698
Assessing Claude Mythos Preview's cybersecurity capabilities
Technical details on how we tested Claude Mythos Preview, a model strikingly capable at computer security tasks, and what we found.
1699
Bend 2 and the Vibe-Coding Trap
Bend just serves as a useful example of my general point regarding vibe-coding as it is recent, high-profile, and has aspects that make it easy to use as an example. I don’t know anything about the author’s history with designing languages or if they actually did consider the tradeoffs below and made what I think is a…
1700
1701
Compromised PyTorch-nightly dependency chain between December 25th – December 30
If you installed PyTorch-nightly on Linux via pip between December 25, 2022 and December 30, 2022, please uninstall it and torchtriton immediately, and use the latest nightly binaries (newer than...
1702
Waymo One: The next step on our self-driving journey
Over time, we hope to make Waymo One available to even more members of the public as we add vehicles and drive in more places.
1703
AlphaEvolve: Gemini-powered coding agent scaling impact across fields
Discover how AlphaEvolve optimizes algorithms for genomics, quantum physics, global infrastructure, and more to accelerate scientific progress and solve real-world challenges.
1704
Amazon saves millions by replacing Sonnet 4.6 with Strands Decider
Two Amazon engineers swapped Claude Sonnet 4.6 for a decision model in one of the world's biggest internal ticketing systems.
1705
Algorithm recovers speech from a potato-chip bag filmed through glass (2014)
Algorithm recovers speech from the vibrations of a potato-chip bag filmed through soundproof glass.
1706
DiffusionGemma: 4x Faster Text Generation
An overview of DiffusionGemma, an exceptionally fast text generation model with up to 4x faster speeds.
1707
1708
The text in Claude Code’s “Extended Thinking” output
Claude Code records each session to disk. Those logs include “thinking blocks” — the model’s own reasoning as it works.
1709
OpenAI models are coming to Google Cloud
OpenAI and Google Cloud are bringing frontier models and Codex workflows to Gemini’s Model Garden, available through Google Cloud Marketplace.
1710