AI300
  1. Kevin Murphy / DeepMind Research Scientist
    1591.
    I agree that for agents (as opposed to making media slop for human consumption), symbolic world models that abstract away from pixels is key! I have a little paper on this (https://t.co/FELmrLYYKl) with @topwasu , @ellisk_kellis and @WLehrach.›
    LLM score 75 · 5 months ago
  2. Clive Chan / OpenAI Hardware
    1592.
    this is incredibly cool https://t.co/3T2oRhdPyU
    LLM score 10 · 5 months ago
  3. Sherwin Wu / OpenAI API, Head of Engineering
    1593.
    Can't remember the last time I've been this excited to listen to a podcast – so excited! Can't come out soon enough.›
    LLM score 25 · 5 months ago
  4. Sherwin Wu / OpenAI API, Head of Engineering
    1594.
    GPT-5.4 is the best model we've ever evaluated on our "Investment Banking" benchmark – it's basically starting to saturate the eval at 87.3%.›
    LLM score 65 · 5 months ago
  5. Aidan McLaughlin / OpenAI Research Scientist
    1595.
    every 6 weeks i realize i’ve been substantially under-ambitious with current models and feel like an idiot
    LLM score 75 · 5 months ago
  6. Andrew Lampinen / Research Scientist at DeepMind
    1596.
    Short post on what I call the "no-magic approach to understanding intelligent systems" — the philosophy I think of as motivating our work on understanding intelligence without resorting to magical thinking about AI or humans! Link below: https://t.co/ew4fhUUguU
    LLM score 75 · 5 months ago
  7. Andrej Karpathy / AI researcher
    1597.
    I packaged up the "autoresearch" project into a new self-contained minimal repo if people would like to play over the weekend.›
    LLM score 85 · 5 months ago
  8. Jeff Dean / Chief Scientist at DeepMind
    1598.
    I'm looking forward to a great discussion with Bill Dally at @nvidia 's GTC event on March 18! https://t.co/rV7ITmrsnQ
    LLM score 20 · 5 months ago
  9. Victoria Lin / Thinking Machines Researcher
    1599.
    ☕ Society will reward tremendously those who can effortlessly spot mistakes made by autonomous agents.
    LLM score 75 · 5 months ago
  10. Boris Cherny / Creator of Claude Code
    1600.
    Released today: /loop /loop is a powerful new way to schedule recurring tasks, for up to 3 days at a time›
    LLM score 55 · 5 months ago
  11. Mark Chen / OpenAI Chief Research Officer
    1601.
    If you give GPT-5.4 a raw dump of the GPT-2 weights and ask for a <5000 byte C program to inference it, GPT-5.4 succeeds in under 15 minutes!›
    LLM score 85 · 5 months ago
  12. Jeff Dean / Chief Scientist at DeepMind
    1602.
    We've been working on the Waxal dataset project since 2021, aiming to enhance the amount of data available for African languages.›
    LLM score 85 · 5 months ago
  13. Soumith Chintala / Thinking Machines CTO
    1603.
    Qwen 3.5 is🔥, and now it's available on 🚀! https://t.co/oXR9RdHdaC
    LLM score 30 · 5 months ago
  14. Merve Noyan / Hugging Face ML Engineer
    1604.
    distillation-attack-as-a-service 🫡 https://t.co/ja4dzxwiKU
    LLM score 20 · 5 months ago
  15. Thomas Wolf / Hugging Face Cofounder
    1605.
    the attack surface keeps increasing https://t.co/e5IT1305OQ
    LLM score 70 · 5 months ago
  16. Kevin Murphy / DeepMind Research Scientist
    1606.
    I am delighted to see a new version of the book by @_sdbuchanan, @druv_pai , @pengwang2003 and @YiMaTweets .›
    LLM score 75 · 5 months ago
  17. Cyris Kissane / Researcher at Flapping Airplanes
    1607.
    The singularity will be bottlenecked by experimentation.
    LLM score 75 · 5 months ago
  18. Clive Chan / OpenAI Hardware
    1608.
    FA4 is out!!!! https://t.co/ChKwKKY9oz
    LLM score 65 · 5 months ago
  19. Alex Zhang / MIT CSAIL PhD
    1609.
    $1,100,000 in cash to write GPU kernels for the latest @GPU_MODE competition holy fuck bro https://t.co/1PczNuIyaV
    LLM score 15 · 5 months ago
  20. Merve Noyan / Hugging Face ML Engineer
    1610.
    wow https://t.co/1fPJMNQkdq
    LLM score 10 · 5 months ago
  21. Lewis Tunstall / Hugging Face Researcher
    1611.
    Democratising distillation attacks, one dataset at a time https://t.co/nNq2Q5wjdm
    LLM score 65 · 5 months ago
  22. Jonathan Ross / TPU Creator
    1612.
    How will things change if for the first time in human history being smart is a commodity?
    LLM score 75 · 5 months ago
  23. Sayak Paul / Hugging Face Researcher
    1613.
    Introducing Modular Diffusers 🔥 The `DiffusionPipeline` abstraction in Diffusers has established a standard in the community.›
    LLM score 75 · 5 months ago
  24. Ben Burtenshaw / Hugging Face Researcher
    1614.
    in SF this weekend for the agentic RL hack with @PyTorch and @UnslothAI .›
    LLM score 20 · 5 months ago
  25. Sayak Paul / Hugging Face Researcher
    1615.
    Yesterday's Diffusers release featured everyone's hot favorite RAEs from @boyangzheng_ et al.›
    LLM score 75 · 5 months ago
  26. Sherwin Wu / OpenAI API, Head of Engineering
    1616.
    Yeah – just like how GPT-5.4 in Codex is fantastic at understanding big, complex codebases, it also thrives in understanding large, complicated Excel files.›
    LLM score 55 · 5 months ago
  27. Andrew White / Edison Scientific Cofounder
    1617.
    68 minutes with GPT-5.4 Pro. It really thinks hard https://t.co/FzsHwnBtm0
    LLM score 20 · 5 months ago
  28. Aidan McLaughlin / OpenAI Research Scientist
    1618.
    research has really been cooking with gas lately.›
    LLM score 35 · 5 months ago
  29. Jack Clark / Anthropic Cofounder
    1619.
    We are still so, so early. https://t.co/i5P5m3ENUP
    LLM score 35 · 5 months ago
  30. Andrej Karpathy / AI researcher
    1620.
    nanochat now trains GPT-2 capability model in just 2 hours on a single 8XH100 node (down from ~3 hours 1 month ago).›
    LLM score 85 · 5 months ago
More