AI300
  1. Sholto Douglas / Researcher at Anthropic
    1651.
    Thais is a good friend and absolute killer - go help her end slop https://t.co/b60ieXK3OQ
    LLM score 25 · 5 months ago
  2. Merve Noyan / Hugging Face ML Engineer
    1652.
    this is almost 1-1 same as my local setup! I use llama server + Pi + Gemma-4 (interchangeably with Qwen3.6) https://t.co/iU7o8HPFxR
    LLM score 65 · 5 months ago
  3. Ben Burtenshaw / Hugging Face Researcher
    1653.
    ICYMI, your agents can one-shot massive data-pipelines with sandboxes and petabytes of storage.›
    LLM score 72 · 5 months ago
  4. Merve Noyan / Hugging Face ML Engineer
    1654.
    this is going to be an eventful week in open-source AI, tune in 👀
    LLM score 15 · 5 months ago
  5. Merve Noyan / Hugging Face ML Engineer
    1655.
    my first finding with this pipeline is that it works well but (rarely) there's a false positive tendency, I don't pass bboxes as tokens but rather overlaid masks/bboxes to judges›
    LLM score 72 · 5 months ago
  6. Cameron Wolfe / Researcher at Netflix
    1656.
    There are three primary grading strategies we can use when evaluating an LLM agent…›
    LLM score 72 · 5 months ago
  7. Alex Zhang / MIT CSAIL PhD
    1657.
    Good harness designs can get around extreme token costs when information is structured.›
    LLM score 72 · 5 months ago
  8. Jeff Dean / Chief Scientist at DeepMind
    1658.
    A good essay by @pgasawa and @profjoeyg on a more nuanced view of AI advances.›
    LLM score 65 · 5 months ago
  9. Dan Fu / VP of Kernels at Together
    1659.
    This is a crazy demo, congrats @krandiash @_albertgu @cartesia on the launch!! https://t.co/U8WClKOS3o
    LLM score 25 · 5 months ago
  10. John Carmack
    1660.
    In the tradition of @fabynou 's game engine books, Bas Smits (on X?) has made the comprehensive book on the Commander Keen games.›
    LLM score 15 · 5 months ago
  11. Albert Gu / Cartesia Chief Scientist
    1661.
    Within the span of a week, we launched streaming TTS (text-to-speech) and STT (speech-to-text) models that topped the leaderboards.›
    LLM score 35 · 5 months ago
  12. Jonas Geiping / AI Researcher in Tübingen
    1662.
    Make your models safer with this one weird trick: ›
    LLM score 72 · 5 months ago
  13. Zixuan Li / Lead Z.ai
    1663.
    GLM-5.2 hasn't officially launched yet (we'll have a formal launch, as usual).›
    LLM score 25 · 5 months ago
  14. Ben Burtenshaw / Hugging Face Researcher
    1664.
    we need to get real and move fast about training our own agents.›
    LLM score 72 · 5 months ago
  15. Ben Burtenshaw / Hugging Face Researcher
    1665.
    colab is so back! https://t.co/vj7uGAEesD
    LLM score 15 · 5 months ago
  16. Sayak Paul / Hugging Face Researcher
    1666.
    Kernels can be instructed to replace the forward passes of compatible layers with their optimized variants 😉›
    LLM score 65 · 5 months ago
  17. Ben Burtenshaw / Hugging Face Researcher
    1667.
    PSA: @satyanadella is telling you to build RL environments! https://t.co/2bNhiwBRAY https://t.co/4cFx88535S
    LLM score 25 · 5 months ago
  18. Jie Tang / Z.ai Cofounder
    1668.
    Z Code! https://t.co/D3VlpuGLQL
    LLM score 25 · 5 months ago
  19. Sayak Paul / Hugging Face Researcher
    1669.
    A tremendously important and short (2mins) read: https://t.co/jiBlax5YoW https://t.co/vNnQRZwHtI
    LLM score 15 · 5 months ago
  20. Jerry Tworek / ex OpenAI VP of RL
    1670.
    Im waiting now for their opinion in muon vs shampoo debate https://t.co/4x6Ty9McWy
    LLM score 15 · 5 months ago
  21. Ben Burtenshaw / Hugging Face Researcher
    1671.
    respectfully handled @NexEcosystem . I hope the city council of Rio continue to use open models, just attribute the work fairly -- especially the leaner shops.›
    LLM score 35 · 5 months ago
  22. Ronak Malde / DeepMind Researcher
    1672.
    “Companies need to turn their workflows, domain knowledge, and accumulated judgment into AI systems that improve with each use.”›
    LLM score 65 · 5 months ago
  23. Devansh Pandey / Standard Intelligence Cofounder
    1673.
    i've wanted a competent version of this for so long.›
    LLM score 25 · 5 months ago
  24. Merve Noyan / Hugging Face ML Engineer
    1674.
    icymi I do photography and I love applying various styles with LoRAs and nano banana on them›
    LLM score 15 · 5 months ago
  25. Merve Noyan / Hugging Face ML Engineer
    1675.
    even though I don't like European way of doing things I kinda dig Mistral's approach to not build gigantic sota generalists but rather focus on medium sized models + fine-tuning-as-a-service for domains›
    LLM score 65 · 5 months ago
  26. Merve Noyan / Hugging Face ML Engineer
    1676.
    ensembling is so back? https://t.co/w76QEq5Spa
    LLM score 35 · 5 months ago
  27. Sherwin Wu / OpenAI API, Head of Engineering
    1677.
    Two weeks with the new Model 3 and … yeah the likelihood of my kids needing to learn how to drive is basically 0% now›
    LLM score 15 · 5 months ago
  28. Jerry Tworek / ex OpenAI VP of RL
    1678.
    Relationship between Anthropic and USG is something to be studied.›
    LLM score 65 · 5 months ago
  29. Leandro von Werra / Hugging Face Head of Research
    1679.
    as "ally" becomes an increasingly volatile term the best outcome is a multipolar world where many companies/countries have frontier models and access can't easily be cut off on a whim https://t.co/Aq8OBmOqXi
    LLM score 45 · 5 months ago
  30. Jerry Tworek / ex OpenAI VP of RL
    1680.
    Decompose and precondition. See the inner beauty in matrices and continuously ascend in the subspace of knowledge
    LLM score 15 · 5 months ago
More