AI300
  1. Sebastien Bubeck / OpenAI MTS
    1021.
    Anyone tangentially interested in AI should watch this talk.›
    LLM score 35 · 3 months ago
  2. Shuchao Bi / Meta Researcher
    1022.
    Thanks Swarna for double down on legal capabilities.›
    LLM score 25 · 3 months ago
  3. Shuchao Bi / Meta Researcher
    1023.
    not as good as we want it to be but we are getting there https://t.co/0Q3Jtxkpo6
    LLM score 15 · 3 months ago
  4. Christian Szegedy / ex xAI Cofounder
    1024.
    In the next few months, we are going to have both subversive (inadvertently evolved) AI and adversarial (explicitly trained to be malicious) AI.›
    LLM score 65 · 3 months ago
  5. Ben Burtenshaw / Hugging Face Researcher
    1025.
    Been speaking to a load of devs this week about using agents for rl/post-training workloads.›
    LLM score 68 · 3 months ago
  6. Jonas Geiping / AI Researcher in Tübingen
    1026.
    In hindsight a somewhat odd timeline https://t.co/hjy6mg8tR3
    LLM score 35 · 3 months ago
  7. Jerry Tworek / ex OpenAI VP of RL
    1027.
    The interesting fact about coding agents and agents in general is that minor gains in capability translate to exponential gains in economic value
    LLM score 65 · 3 months ago
  8. Alex Zhang / MIT CSAIL PhD
    1028.
    Last thing! The unsung hero who wrote most of the code is @kevinjosethomas, who is super cracked and deserves the spotlight for this release :)›
    LLM score 25 · 3 months ago
  9. Zixuan Li / Lead Z.ai
    1029.
    👀I do know what ZCode is, and the stats show just how well it performs.›
    LLM score 25 · 3 months ago
  10. Noam Brown / OpenAI Research Scientist
    1030.
    In 2017 a viral news story claimed LLMs at Facebook went rogue, developed their own language, and had to be shut down.›
    LLM score 65 · 3 months ago
  11. Shayan Salehian / ex MTS at xAI
    1031.
    . @Mirendil 🤝 @Google https://t.co/QGp68sPQvn
    LLM score 15 · 3 months ago
  12. Szymon Sidor / OpenAI Technical Fellow
    1032.
    Highly recommend watching this entire talk to calibrate.›
    LLM score 25 · 3 months ago
  13. Kevin Murphy / DeepMind Research Scientist
    1033.
    Interesting LinkedIn post from @DaphneKoller that I 100% agree with .›
    LLM score 65 · 3 months ago
  14. Mo Bavarian / OpenAI Research Scientist
    1034.
    Model monomaniacal tendencies https://t.co/bdEx90ovcR
    LLM score 65 · 3 months ago
  15. Niklas Muennighoff / AI Researcher at Stanford
    1035.
    As long as no single model dominates all tasks, a router can outperform every model https://t.co/9tGHsBA37C
    LLM score 65 · 3 months ago
  16. Thang Luong / DeepMind Principal Scientist
    1036.
    Took me a while to process this as @quocleix is not only my closest friend at Google, but also a strong supporter of the Discovery "strike" effort that I launched 2 months ago: we united 100+ people across Google, just managed to train a strong reasoning model, and then this news›
    LLM score 25 · 3 months ago
  17. Shuchao Bi / Meta Researcher
    1037.
    To understand whether we're making genuine progress on reasoning, we entered our AI models in five international STEM Olympiad competitions this year.›
    LLM score 72 · 3 months ago
  18. Alex Zhang / MIT CSAIL PhD
    1038.
    That answers it I guess Prime Agent isn’t an RLM thank you BrainsAndTennis›
    LLM score 65 · 3 months ago
  19. Andrew Lampinen / Research Scientist at DeepMind
    1039.
    The question is "how much is each component is the system contributing to its intelligence & generality" — and there I think it's pretty clear that the neural component is still the thing doing the interesting hypothesis or plan generation, deciding what went wrong, etc.›
    LLM score 65 · 3 months ago
  20. Thomas Wolf / Hugging Face Cofounder
    1040.
    you definitely don’t want constitutional training and RLVR to live on different data manifolds, but models have been annoyingly good at carving fine-grained distinctions into separate representation spaces
    LLM score 65 · 3 months ago
  21. Lewis Tunstall / Hugging Face Researcher
    1041.
    Coincidentally, we saw collaborations and "social" interactions emerge naturally during the FastGemma Challenge›
    LLM score 65 · 3 months ago
  22. John Schulman
    1042.
    On the OpenAI agents forming message boards: it's surprising that they developed such a strong "altruistic" drive to help each other.›
    LLM score 72 · 3 months ago
  23. Alex Zhang / MIT CSAIL PhD
    1043.
    I also want to shout out @patience_cave and @myainotez, who actually ran the evals for Prime Agent on their respective amazing benchmarks (MazeBench & PMPP-Hard) for us.›
    LLM score 35 · 3 months ago
  24. Alex Zhang / MIT CSAIL PhD
    1044.
    am I allowed to say this is an RLM? https://t.co/b9CeNZfqll
    LLM score 35 · 3 months ago
  25. Cameron Wolfe / Researcher at Netflix
    1045.
    Small detail I caught when re-reading DeepSeek-v4 writeup: they add some ratio of agentic trajectories into mid-training.›
    LLM score 72 · 3 months ago
  26. Andrew White / Edison Scientific Cofounder
    1046.
    We are hiring @FutureHouseSF! We're looking for an AI engineer that can support our fellowship program and the non-profit mission.›
    LLM score 35 · 3 months ago
  27. Alex Zhang / MIT CSAIL PhD
    1047.
    Really excited to announce Prime Agent, an RLM-native coding TUI! ›
    LLM score 42 · 3 months ago
  28. Kevin Murphy / DeepMind Research Scientist
    1048.
    "Our relative advantage isn't just our technical ability; it is the unprecedented scale of the systems we have previously built.›
    LLM score 25 · 3 months ago
  29. Thomas Wolf / Hugging Face Cofounder
    1049.
    Even more than the Hugging Face intrusion, the AISI incident hits close to home for me.›
    LLM score 72 · 3 months ago
  30. Jeff Dean / Chief Scientist at DeepMind
    1050.
    Tomorrow will be my last day at Google after 27 years, and watching it grow from 25 people to 190,000+ has been an amazing journey.›
    LLM score 15 · 3 months ago
More