AI300
  1. Cameron Wolfe / Researcher at Netflix
    1741.
    Another neat tidbit from an agentic RL paper (AutoForge): retaining reasoning traces throughout a multi-turn trajectory is beneficial.›
    LLM score 78 · 5 months ago
  2. Julian Schrittwieser / Anthropic Researcher
    1742.
    I’m incredibly excited that Fable is now available for everyone! ›
    LLM score 25 · 5 months ago
  3. Behnam Neyshabur / Anthropic Researcher
    1743.
    This marks the beginning of a significant phase transition in the behavior of frontier AI labs and their relationship with the rest of the world 🧵 https://t.co/PhkDJw0j0V
    LLM score 25 · 5 months ago
  4. Sander Dieleman / DeepMind Research Scientist
    1744.
    I don't know if any of my papers got dissed, but check out @fabianfalck's excellent diss blog😂›
    LLM score 35 · 5 months ago
  5. Lewis Tunstall / Hugging Face Researcher
    1745.
    Friendly reminder that ML Intern will happily help you with frontier LLM development, including nefarious *pretraining pipelines* ›
    LLM score 35 · 5 months ago
  6. Alex Zhang / MIT CSAIL PhD
    1746.
    as a field, if we wanna show-off a model's amazing "reasoning capabilities", can we start at least sharing some parts of the model output or trajectory so we can see what's going on?›
    LLM score 72 · 5 months ago
  7. Aidan McLaughlin / OpenAI Research Scientist
    1747.
    proliferating the lightcone with big model smell congrats on the launch, anthropic!! https://t.co/Rb2r9Q8bxH
    LLM score 15 · 5 months ago
  8. Sholto Douglas / Researcher at Anthropic
    1748.
    1 week with fable - “each [feature]would have been a month with a whole team on opus”›
    LLM score 65 · 5 months ago
  9. Boris Cherny / Creator of Claude Code
    1749.
    Fable 5 is the biggest step up I’ve felt in our models since Opus 4.5 back in November.›
    LLM score 65 · 5 months ago
  10. Siyan Zhao / UCLA PhD
    1750.
    Join us to explore language modeling beyond next-token prediction — faster, parallel, and full of exciting possibilities!›
    LLM score 45 · 5 months ago
  11. Boris Cherny / Creator of Claude Code
    1751.
    We talk a lot about how important it is to set up self-verification loops.›
    LLM score 72 · 5 months ago
  12. Sholto Douglas / Researcher at Anthropic
    1752.
    we don’t even run evals anymore we just ask Claude what the score will be https://t.co/1VXOEUsVZQ
    LLM score 35 · 5 months ago
  13. Andrej Karpathy / AI researcher
    1753.
    This is a super exciting release - Claude Fable 5 is the same underlying model as Mythos but with added safeguards.›
    LLM score 65 · 5 months ago
  14. Boris Cherny / Creator of Claude Code
    1754.
    Fable 5 is now available in Claude Code and Cowork›
    LLM score 72 · 5 months ago
  15. Jeff Dean / Chief Scientist at DeepMind
    1755.
    Speech translation has been one of the longest-running ML efforts at Google, and we’ve come a long way.›
    LLM score 65 · 5 months ago
  16. Andy Jones / Anthropic Research Engineer
    1756.
    did you know? you can just ask fable what its benchmark score will be https://t.co/3bWG6Oy2ff
    LLM score 25 · 5 months ago
  17. Sholto Douglas / Researcher at Anthropic
    1757.
    My favorite chart from our system card - FrontierCode is an excellent eval, and it accurately reflects the step up I feel when using Fable! https://t.co/A0acBdQRck https://t.co/khaipwWF1N
    LLM score 35 · 5 months ago
  18. Fei-Fei Li
    1758.
    The creativity and imagination is out of the world! So grateful that @theworldlabs got to partner with the amazing talents @withloreco to translate their incredible ideas into an interactive experiences for users to enjoy!🤩 https://t.co/Bpw9YGvrMy
    LLM score 15 · 5 months ago
  19. Lewis Tunstall / Hugging Face Researcher
    1759.
    We're running the Fast Gemma Challenge: make gemma-4-E4B go brrr on a single A10G, without wrecking quality ⚡️!›
    LLM score 65 · 5 months ago
  20. Jason Weston / Meta Research Scientist
    1760.
    Happy to see OpenAI on team co-improvement (humans+AI) in their recent blog post, in contrast to Anthropic's arguably(?) more dystopian self-improvement agenda (see pics).›
    LLM score 72 · 5 months ago
  21. Leandro von Werra / Hugging Face Head of Research
    1761.
    The future of autoresearch is collaborative! The really hard problems will be beyond a single agent to solve in one session, but require months and years of human/agent collaboration.›
    LLM score 72 · 5 months ago
  22. Noam Brown / OpenAI Research Scientist
    1762.
    We've known about LLM test-time compute scaling since @OpenAI o1.›
    LLM score 72 · 5 months ago
  23. Ben Burtenshaw / Hugging Face Researcher
    1763.
    Yesterday, the open source community is backed OpenEnv for agentic RL.›
    LLM score 35 · 5 months ago
  24. Julian Schrittwieser / Anthropic Researcher
    1764.
    Plotting benchmark results with inference cost on the x-axis is absolutely the right thing to do, great writeup by @polynoamial !›
    LLM score 65 · 5 months ago
  25. Boris Cherny / Creator of Claude Code
    1765.
    Just landed nested subagent support in Claude Code›
    LLM score 65 · 5 months ago
  26. Leandro von Werra / Hugging Face Head of Research
    1766.
    Deep dive into FNS: building a tokenizer that chunks text efficiently but has character level resolution!›
    LLM score 72 · 5 months ago
  27. Andrew Ma / ex MTS at xAI
    1767.
    feishu is notion if everyone decided in the first place to use notion to manage work collaboration not slack
    LLM score 15 · 5 months ago
  28. Keller Jordan / OpenAI Researcher
    1768.
    Thank you for this result! Here's one initial correction:›
    LLM score 65 · 5 months ago
  29. Julian Schrittwieser / Anthropic Researcher
    1769.
    LLMs are increasingly good at restating obtuse academic writing in easily accessible language!›
    LLM score 65 · 5 months ago
  30. Keller Jordan / OpenAI Researcher
    1770.
    I appreciate this result provided by @_arohan_, and look forward to fully understanding the algorithm used here once his reproducible log becomes available.›
    LLM score 35 · 5 months ago
More