AI300
  1. John Carmack
    1.
    Paper review: LeWorldModel: Stable End-to-End Joint-Embedding Predictive Architecture from Pixels›
    LLM score 92 · about 3 hours ago
  2. Eric Jang / ex VP of AI at 1X Robotics
    2.
    There is currently a lot of global anxiety around AI, what things will be meaningful to work on, and how to make it turn out well for all people.›
    LLM score 85 · about 4 hours ago
  3. Merve Noyan / Hugging Face ML Engineer
    3.
    IBM just dropped Granite 4.0-3B-Vision, new vision language model for documents›
    LLM score 75 · about 5 hours ago
  4. Cameron Wolfe / Researcher at Netflix
    4.
    I just published a blog that covers 30+ popular LLM evals / benchmarks and how they are created.›
    LLM score 92 · about 5 hours ago
  5. Leandro von Werra / Hugging Face Head of Research
    5.
    Six years into making and finally TRL is at v1.0! ›
    LLM score 85 · about 6 hours ago
  6. Ben Burtenshaw / Hugging Face Researcher
    6.
    hugging face spaces is being repurposed as an agentic rl platform.›
    LLM score 92 · about 7 hours ago
  7. Jie Tang / Z.ai Cofounder
    7.
    Thanks for mentioning GLM. We will soon release GLM 5.1 with better performances on many other benchmarks … https://t.co/G18fUYN8zW
    LLM score 75 · about 11 hours ago
  8. Andrej Karpathy / AI researcher
    8.
    New supply chain attack this time for npm axios, the most popular HTTP client library with 300M weekly downloads.›
    LLM score 92 · about 16 hours ago
  9. Alex Zhang / MIT CSAIL PhD
    9.
    Cool paper and generally a good direction, I actually think the RLM paper and this paper are arguing roughly for the same thing despite it not appearing that way on paper (i.e.›
    LLM score 85 · about 17 hours ago
  10. Mark Chen / OpenAI Chief Research Officer
    10.
    Really proud of how our auto compaction turned out.›
    LLM score 92 · about 18 hours ago
  11. Chelsea Finn / Physical Intelligence Cofounder
    11.
    Agents like Claude Code do amazing things relying critically on hand-engineered harnesses›
    LLM score 95 · about 22 hours ago
  12. Tri Dao / Chief Scientist at Together
    12.
    It's my favorite kind of work: linear algebra insight + fast kernels.›
    LLM score 92 · 1 day ago
  13. Jason Weston / Meta Research Scientist
    13.
    🔗Learning to Aggregate through Online RL🎯 ParaGator🔀🐊: strong parallel reasoning aggregation›
    LLM score 92 · 1 day ago
  14. Thomas Wolf / Hugging Face Cofounder
    14.
    the LLM is the computer https://t.co/3fLyn1VPI7
    LLM score 85 · 1 day ago
  15. Ronak Malde / DeepMind Researcher
    15.
    I have long felt that agent harnesses - even claude code - are too restrictive, because they are still designed by humans.›
    LLM score 92 · 2 days ago
  16. Cameron Wolfe / Researcher at Netflix
    16.
    New blog post coming out tomorrow morning on LLM benchmarking.›
    LLM score 92 · 2 days ago
  17. Andrew Lampinen / Research Scientist at DeepMind
    17.
    Important point, and definitely something to check for when creating a vision-language benchmark! (I thought I saw a paper arguing for just that a half-year or so ago but can't find it now, so maybe I hallucinated.) This also reminds me of some of the kinds of confabulation 1/2›
    LLM score 85 · 2 days ago
  18. Ben Burtenshaw / Hugging Face Researcher
    18.
    gastown on @karpathy's autoresearch is just the wildest ride.›
    LLM score 92 · 3 days ago
  19. Sergey Levine / Physical Intelligence Cofounder
    19.
    Didn't think that "drones with grippers" was in the cards for a likely embodiment for pi models, but there we have it.›
    LLM score 92 · 3 days ago
  20. Karol Hausman / Physical Intelligence Cofounder
    20.
    breathe intelligence into everything that moves 🦾 ›
    LLM score 65 · 3 days ago
  21. Andrej Karpathy / AI researcher
    21.
    - Drafted a blog post - Used an LLM to meticulously improve the argument over 4 hours.›
    LLM score 92 · 3 days ago
  22. Alex Zhang / MIT CSAIL PhD
    22.
    sorry I know this is an educational post but I just thought the hook was so funny›
    LLM score 75 · 3 days ago
  23. Jonathan Ross / TPU Creator
    23.
    The scaling laws paper is the most misread paper in AI.›
    LLM score 92 · 3 days ago
  24. Ben Burtenshaw / Hugging Face Researcher
    24.
    genuine question: for everyone that thinks CLI beats MCP, how do you authenticate services from chat apps using sandboxes?
    LLM score 92 · 4 days ago
  25. Adnan Esmail / Physical Intelligence Cofounder
    25.
    Spent years building drone systems the hard way - custom controllers, hand-tuned dynamics, endless edge cases.›
    LLM score 92 · 4 days ago
  26. Clive Chan / OpenAI Hardware
    26.
    same problem tbh, you’re laying out a bunch of tightly interconnected stuff in 2D https://t.co/Q5tiPMWd7m
    LLM score 75 · 4 days ago
  27. Andrew Ma / ex MTS at xAI
    27.
    We all know biology is a computational substrate but does anyone have ideas for how to make viable biological computers that are not brains
    LLM score 92 · 4 days ago
  28. Cameron Wolfe / Researcher at Netflix
    28.
    I really like the strategy used by CursorBench to evaluate Composer 2.›
    LLM score 92 · 4 days ago
  29. Merve Noyan / Hugging Face ML Engineer
    29.
    fav papers from 3DV #3 DetailGen3D refines geometric texture of 3D assets, model learns to do coarse-to-fine transformation›
    LLM score 92 · 4 days ago
  30. Sayak Paul / Hugging Face Researcher
    30.
    The more I learn to better pair code with Claude, the more I realize the importance of solid foundations and gold-standard references.›
    LLM score 85 · 4 days ago
More