AI300
  1. Demis Hassabis / CEO of DeepMind
    3541.
    First LLM contact from space 🛰️ using our highly efficient open source Gemma models! Huge congrats to @PhilipJohnston and the @Starcloud_Inc_ team! https://t.co/JFuh9Y8a1f
    LLM score 20 · 11 months ago
  2. Noam Brown / OpenAI Research Scientist
    3542.
    An important lesson that ARC-AGI has internalized, but not many others have, is that benchmark perf is a function of test-time compute.›
    LLM score 85 · 11 months ago
  3. Noam Brown / OpenAI Research Scientist
    3543.
    IMO GDPVal is the most important result from our @OpenAI GPT-5.2 launch.›
    LLM score 80 · 11 months ago
  4. Noam Brown / OpenAI Research Scientist
    3544.
    I'm also really happy that @OpenAI was willing to publish the original GDPVal results showing Claude ahead of ChatGPT.›
    LLM score 60 · 11 months ago
  5. Demis Hassabis / CEO of DeepMind
    3545.
    The UK is an amazing place for science & innovation.›
    LLM score 50 · 11 months ago
  6. Demis Hassabis / CEO of DeepMind
    3546.
    We’re also announcing a new partnership with @AISecurityInst that builds on two years working together & will focus on foundational safety and security research essential for realising AI’s potential to benefit humanity.›
    LLM score 45 · 11 months ago
  7. Demis Hassabis / CEO of DeepMind
    3547.
    I’ve always believed that AI will be the most transformational technology of our time, and partnerships like this are vital to turning that potential into real progress @SciTechgovuk.›
    LLM score 20 · 11 months ago
  8. Radhakrishnan Venkataramani / ex MTS at xAI
    3548.
    Now that coding models have got better and going to get even better, the most refreshing change i wish/pray for›
    LLM score 82 · 11 months ago
  9. Andrej Karpathy / AI researcher
    3549.
    Quick new post: Auto-grading decade-old Hacker News discussions with hindsight›
    LLM score 80 · 11 months ago
  10. Tim Dettmers / Research Scientist at Ai2
    3550.
    Many people think AI will continue improve towards AGI.›
    LLM score 80 · 11 months ago
  11. Tim Dettmers / Research Scientist at Ai2
    3551.
    My new blog post discusses the physical reality of computation and why this means we will not see AGI or any meaningful superintelligence: https://t.co/jsAKQ6T3gC
    LLM score 80 · 11 months ago
  12. Andrej Karpathy / AI researcher
    3552.
    In today's episode of programming horror... In the Python docs of random.seed() def, we're told›
    LLM score 40 · 11 months ago
  13. Noam Brown / OpenAI Research Scientist
    3553.
    From inception to release, the journal publication process can easily take over a year.›
    LLM score 70 · 11 months ago
  14. Igor Babuschkin / Cofounder of xAI
    3554.
    SGLang is the best inference framework for LLMs.›
    LLM score 20 · 11 months ago
  15. Andrej Karpathy / AI researcher
    3555.
    @ChrSzegedy I could certainly imagine that "nesting" the simulation might be too "effortful" for the model, compute or data density wise.›
    LLM score 60 · 11 months ago
  16. Noam Brown / OpenAI Research Scientist
    3556.
    @Thom_Wolf Putnam is more knowledge-based whereas IMO requires more creativity and more time per problem, so Putnam is easier for LLMs.
    LLM score 70 · 11 months ago
  17. Andrej Karpathy / AI researcher
    3557.
    @DimitrisPapail There is definitely work going into engineering the "you" simulation - the personality that gets all the rewards in verifiable problems, or all the upvotes from users/judge LLMs, or mimics the responses of SFT, and there is an emergent composite personality from that.›
    LLM score 30 · 11 months ago
  18. Andrej Karpathy / AI researcher
    3558.
    Don't think of LLMs as entities but as simulators.›
    LLM score 85 · 11 months ago
  19. Demis Hassabis / CEO of DeepMind
    3559.
    Gemini has always had exceptionally strong multimodal capabilities.›
    LLM score 15 · 11 months ago
  20. Leo Gao / OpenAI Alignment Researcher
    3560.
    New post: An Ambitious Vision for Interpretability›
    LLM score 65 · 11 months ago
  21. Noam Brown / OpenAI Research Scientist
    3561.
    @deredleritt3r Ah yeah that could have been worded better.›
    LLM score 75 · 11 months ago
  22. Noam Brown / OpenAI Research Scientist
    3562.
    @deredleritt3r We didn’t use the same exact model for the IMO, IOI, and ICPC.›
    LLM score 70 · 11 months ago
  23. Demis Hassabis / CEO of DeepMind
    3563.
    Gemini 3 Deep Think is now available for Google AI Ultra subscribers in the @GeminiApp, incorporating our gold medal winning IMO and ICPC technologies! 🏅With its parallel thinking capabilities it can tackle highly complex maths & science problems - enjoy! https://t.co/5BXWmYyor6
    LLM score 35 · 11 months ago
  24. Noam Shazeer
    3564.
    Gemini 3 Deep Think mode is live for Ultra users today.›
    LLM score 75 · 11 months ago
  25. Quoc Le / Google Fellow
    3565.
    Gemini 3 Deep Think mode is now live in the Gemini app for Ultra users.›
    LLM score 75 · 11 months ago
  26. Radhakrishnan Venkataramani / ex MTS at xAI
    3566.
    Fascinating. Coding models are going to have step function improvements once›
    LLM score 85 · 11 months ago
  27. Mark Chen / OpenAI Chief Research Officer
    3567.
    GPT-5 generated the key insight for a paper accepted to Physics Letters B, a serious and reputable peer-reviewed journal.›
    LLM score 85 · 11 months ago
  28. Noam Brown / OpenAI Research Scientist
    3568.
    @fchollet Do you consider any humans to have an understanding of differential equations?
    LLM score 40 · 11 months ago
  29. Guangxuan Xiao / Thinking Machines Researcher
    3569.
    It turns out that reducing the range of FP4 values is actually a key to unlocking stable low-bit training 😆 https://t.co/vle4zeELDo
    LLM score 80 · 11 months ago
  30. Kevin Murphy / DeepMind Research Scientist
    3570.
    This is a cool Julia version of my Jax library for Bayesian Structural Time Series modeling (https://t.co/T5gQOQPN9o) from the folks at @ReactiveBayes.›
    LLM score 75 · 11 months ago
More