AI300
  1. Sherwin Wu / OpenAI API, Head of Engineering
    1321.
    Definitely want you building on top of our products! https://t.co/SX5nkJmSfV
    LLM score 85 · 4 months ago
  2. Eric Jang / ex VP of AI at 1X Robotics
    1322.
    Pretty fun puzzle generator for alignment and safety researchers https://t.co/pclOnYRVFT
    LLM score 85 · 4 months ago
  3. Andrej Karpathy / AI researcher
    1323.
    Farzapedia, personal wikipedia of Farza, good example following my Wiki LLM tweet.›
    LLM score 92 · 4 months ago
  4. Andrej Karpathy / AI researcher
    1324.
    Something I've been thinking about - I am bullish on people (empowered by AI) increasing the visibility, legibility and accountability of their governments.›
    LLM score 85 · 4 months ago
  5. Cameron Wolfe / Researcher at Netflix
    1325.
    Gemma 4 uses a few interesting architectural tricks (some of which were also used in Gemma-3n) that I haven't learned about before: Shared KV cache and Per-layer Embeddings (PLE).›
    LLM score 95 · 4 months ago
  6. Damek Davis / Assoc. Professor Wharton Stats
    1326.
    I added this problem to the Optimization Constants in Mathematics repository that I'm maintaining with @PI010101 and Terry Tao.›
    LLM score 92 · 4 months ago
  7. Andrej Karpathy / AI researcher
    1327.
    Wow, this tweet went very viral! I wanted share a possibly slightly improved version of the tweet in an "idea file".›
    LLM score 85 · 4 months ago
  8. Ben Burtenshaw / Hugging Face Researcher
    1328.
    PSA: models on hugging face are free to use.›
    LLM score 92 · 4 months ago
  9. Ben Burtenshaw / Hugging Face Researcher
    1329.
    you have two options to liberate your claw and never look back.›
    LLM score 92 · 4 months ago
  10. Boris Cherny / Creator of Claude Code
    1330.
    Starting tomorrow at 12pm PT, Claude subscriptions will no longer cover usage on third-party tools like OpenClaw.›
    LLM score 85 · 4 months ago
  11. Andrew Lampinen / Research Scientist at DeepMind
    1331.
    When and how can test-time thinking allow models to use information latent in their training data? What are the benefits and tradeoffs relative to other solutions like synthetic data? Pleased to share (after a long delay) an exploration of these issues:
    LLM score 95 · 4 months ago
  12. Sayak Paul / Hugging Face Researcher
    1332.
    We're shipping an elaborate guide on how to profile diffusion pipelines in Diffusers to set them up for success with `torch.compile` 🔥›
    LLM score 92 · 4 months ago
  13. Jonathan Ross / TPU Creator
    1333.
    A single employee with no venture backing just built a $1.8B business using AI tools and $20K.›
    LLM score 85 · 4 months ago
  14. Jerry Tworek / ex OpenAI VP of RL
    1334.
    One thing I found is that silencing conscious verbal thought process often leads to richer multidimensional subconscious processing https://t.co/ykhTzRYIEN
    LLM score 75 · 4 months ago
  15. Jason Weston / Meta Research Scientist
    1335.
    🧮 Reasoning over Mathematical Objects 🧮 Our 70-page(!) paper is out on arXiv, as covered by several of our recent blog posts.›
    LLM score 92 · 4 months ago
  16. Christian Szegedy / ex xAI Cofounder
    1336.
    Papers are the proof of work for academic transactions.
    LLM score 72 · 4 months ago
  17. Andrej Karpathy / AI researcher
    1337.
    LLM Knowledge Bases Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest.›
    LLM score 92 · 4 months ago
  18. Yoshua Bengio
    1338.
    Un tour d’horizon de ma vision pour une IA plus sécuritaire et au service de l'humanité, dans le cadre de la série « les gourous de l’IA » de @lemondefr ›
    LLM score 85 · 4 months ago
  19. Merve Noyan / Hugging Face ML Engineer
    1339.
    run OpenClaw, Hermes Agent and Pi with Gemma 4 with few lines of change 🔥 https://t.co/kvv444Ht7O
    LLM score 85 · 4 months ago
  20. Cameron Wolfe / Researcher at Netflix
    1340.
    Huge fan of TBPN, but not sure how I feel about a frontier company acquiring the biggest media source in AI.›
    LLM score 82 · 4 months ago
  21. Ben Burtenshaw / Hugging Face Researcher
    1341.
    Here's the definitive gemma 4 release day tinker challenge.›
    LLM score 85 · 4 months ago
  22. Jeff Dean / Chief Scientist at DeepMind
    1342.
    Today we're releasing Gemma 4, our new family of open foundation models, built on the same research and technology as our Gemini 3 series.›
    LLM score 15 · 4 months ago
  23. Damek Davis / Assoc. Professor Wharton Stats
    1343.
    Spent last month making a tiny harness around codex, which helped me prove a new result on the following question:›
    LLM score 95 · 4 months ago
  24. Ben Burtenshaw / Hugging Face Researcher
    1344.
    . @googlegemma have open sourced the perfect model for local open source agents.›
    LLM score 82 · 4 months ago
  25. Merve Noyan / Hugging Face ML Engineer
    1345.
    future is local 🔥 Google DeepMind just released Gemma 4: local frontier in many sizes, all modalities with free license 🤯›
    LLM score 92 · 4 months ago
  26. Demis Hassabis / CEO of DeepMind
    1346.
    Excited to launch Gemma 4: the best open models in the world for their respective sizes.›
    LLM score 75 · 4 months ago
  27. Leandro von Werra / Hugging Face Head of Research
    1347.
    Congrats GDM on the Gemma 4 release! The super strong 30B model will be very nice for anyone who wants to fine-tune/mid-train models without the pain of MoEs 💥 https://t.co/MJ8Za7pENW
    LLM score 82 · 4 months ago
  28. John Carmack
    1348.
    Without getting all the way down to performance counters, GPU power from nvidia-smi is a better indicator of true utilization than job scheduling or “gpu busy”.›
    LLM score 92 · 4 months ago
  29. Leandro von Werra / Hugging Face Head of Research
    1349.
    Status update: I have been reinstated under the condition to never allow the usage of `hf-sync` and `hf-mount` again.›
    LLM score 85 · 4 months ago
  30. Ben Burtenshaw / Hugging Face Researcher
    1350.
    a perfect time to open source a few of those amazing rl environments.›
    LLM score 85 · 4 months ago
More