AI300
  1. Andrew Lampinen / Research Scientist at DeepMind
    151.
    Indeed. I'll point to a few papers that suggest why this isn't a problem (and in fact can be beneficial) below: https://t.co/8fJYJ6t67W
    LLM score 65 · 12 days ago
  2. Chelsea Finn / Physical Intelligence Cofounder
    152.
    The Pi API is officially in beta! 🤖 Many people have already been using it to fine-tune π models for their robots -- below are a few›
    LLM score 35 · 13 days ago
  3. Sergey Levine / Physical Intelligence Cofounder
    153.
    Even a year ago I don't think I would have believed that PI models can run on such a huge range of different robots.›
    LLM score 35 · 13 days ago
  4. Simon Zhai / ex MTS at xAI
    154.
    getting emotional damage now https://t.co/fV9KUow2ML
    LLM score 25 · 13 days ago
  5. Andrew White / Edison Scientific Cofounder
    155.
    I feel even more confused on quiet periods for IPOs https://t.co/gs0FxvQLnD
    LLM score 15 · 13 days ago
  6. Andrew White / Edison Scientific Cofounder
    156.
    Excited for datacenters in space so LLMs will be truly extraterrestrial aliens
    LLM score 15 · 13 days ago
  7. Thang Luong / DeepMind Principal Scientist
    157.
    Gemini-4 Argon ranks #5 on Agent @arena.›
    LLM score 25 · 13 days ago
  8. Yoshua Bengio
    158.
    I've had so many insightful conversations over the years with @lessig, an expert on AI regulation and legislation, including more recent ones on the cyber incidents of this past summer.›
    LLM score 25 · 13 days ago
  9. Merve Noyan / Hugging Face ML Engineer
    159.
    people are building coolest stuff with EmbeddingGemma 2, I'm so happy embedding models now get the attention they deserve https://t.co/92nkwQoAkw
    LLM score 45 · 13 days ago
  10. Sholto Douglas / Researcher at Anthropic
    160.
    I’m betting this becomes a textbook case.›
    LLM score 45 · 13 days ago
  11. Leandro von Werra / Hugging Face Head of Research
    161.
    Astonishing: gave Claude (Opus 5.5) a few pens and paper and asked it to make beautiful art for a contest.›
    LLM score 25 · 13 days ago
  12. Keller Jordan / OpenAI Researcher
    162.
    TIL that after Garry Kasparov lost his 1997 match against Deep Blue, he accused the IBM team of having cheated by secretly using human expertise https://t.co/fcBsAr2IoC
    LLM score 25 · 13 days ago
  13. Horace He / Thinking Machines Founding Engineer
    163.
    If training runs on Tinker were fully bitwise deterministic, how valuable would this be to you? Curious about opinions across our user demographic...
    LLM score 35 · 13 days ago
  14. Demis Hassabis / CEO of DeepMind
    164.
    Amazing to see research move from the lab to real-world clinical utility - congrats to the team on this milestone & @TheLancet paper! https://t.co/Jq0G4D6Cx2
    LLM score 35 · 13 days ago
  15. Jerry Tworek / ex OpenAI VP of RL
    165.
    Generalization in 2026: we trained a chonkers on the energy consumption of London and it is good in some things https://t.co/jKmAEP8WzC
    LLM score 25 · 13 days ago
  16. Jack Clark / Anthropic Cofounder
    166.
    I will be giving the Per Jacobsson Foundation Lecture at the IMF conference in Thailand on Saturday, during which I'm going to make the case that we are entering a period of massive, sudden change catalyzed by the advancing capabilities of AI systems.›
    LLM score 15 · 13 days ago
  17. Keller Jordan / OpenAI Researcher
    167.
    Seems like an interesting new speedrun (I don't know whether it's inevitable, but this is a good way to find out.) https://t.co/lfif0RnLxG
    LLM score 35 · 14 days ago
  18. Sholto Douglas / Researcher at Anthropic
    168.
    people have been asking 'if models are so good where is all the new software' for the last year or so.›
    LLM score 65 · 14 days ago
  19. Christian Szegedy / ex xAI Cofounder
    169.
    While the points of this article are technically valid, it is kind of a demonstration how AI can be creative.›
    LLM score 65 · 14 days ago
  20. Siyan Zhao / UCLA PhD
    170.
    Is on-policy sampling always good for distilling any teacher–student pair?›
    LLM score 82 · 14 days ago
  21. Zixuan Li / Lead Z.ai
    171.
    Saw GLM-Legal getting a lot of discussion on HN.›
    LLM score 65 · 14 days ago
  22. Jason Wei / AI Researcher at Meta
    172.
    muse wonder is out! scaling prevails… it has been so fun to be part of the powerful yet nimble ship of Meta, where in just over a year, much of the company pivoted to contribute to the model, ranging from the infra engineers who started building tools for experiment https://t.co/of2NCvxBSX
    LLM score 35 · 14 days ago
  23. Shuchao Bi / Meta Researcher
    173.
    I run every major life decision with our new model now.›
    LLM score 25 · 14 days ago
  24. Thomas Wolf / Hugging Face Cofounder
    174.
    open and state-of-the-art decision models - just saying https://t.co/RAOBOtOfIh
    LLM score 25 · 14 days ago
  25. Zixuan Li / Lead Z.ai
    175.
    GLM-Legal is our first specialized model for professional work.›
    LLM score 35 · 14 days ago
  26. Thomas Wolf / Hugging Face Cofounder
    176.
    Blue = what we knew. Green = what's new.›
    LLM score 35 · 14 days ago
  27. Christian Szegedy / ex xAI Cofounder
    177.
    Worth watching https://t.co/W7QSpYc6IW
    LLM score 15 · 14 days ago
  28. Sander Dieleman / DeepMind Research Scientist
    178.
    Turns out you can use distribution matching distillation (DMD) to turn diffusion/flow models into looped models that iteratively refine samples directly in data space🤯›
    LLM score 82 · 14 days ago
  29. Sander Dieleman / DeepMind Research Scientist
    179.
    This is a beautiful analysis of the order in which diffusion models resolve uncertainty about information at different abstraction levels in different representation spaces, with significant implications for latent diffusion.›
    LLM score 72 · 14 days ago
  30. Leandro von Werra / Hugging Face Head of Research
    180.
    We released Carbon-A: a DNA model that proposed over 500M gene candidates across 22k species!!›
    LLM score 35 · 14 days ago
More