AI300
  1. Ben Burtenshaw / Hugging Face Researcher
    2401.
    here's a hands on guide to setup multi-agent autoresearch by @karpathy.›
    LLM score 92 · 7 months ago
  2. Tim Dettmers / Research Scientist at Ai2
    2402.
    So cool to see that open-source, with open experimentation (and with the help of someone posting blog posts about their personal research), can yield a very robust method for MoE balancing.›
    LLM score 92 · 7 months ago
  3. Ben Burtenshaw / Hugging Face Researcher
    2403.
    yea tokenizers! super insightful thread on how and why a model's tokenizers can be changed without pretaining.›
    LLM score 82 · 7 months ago
  4. Eric Jang / ex VP of AI at 1X Robotics
    2404.
    I enjoyed the Jensen + @dwarkesh_sp podcast.›
    LLM score 75 · 7 months ago
  5. Albert Gu / Cartesia Chief Scientist
    2405.
    a dynamical systems point of view, which looks like an SSM applied along the residual stream, informs more principled ways to scale looped architectures https://t.co/Ry09zXCXY1
    LLM score 92 · 7 months ago
  6. Ronak Malde / DeepMind Researcher
    2406.
    Ultra long horizon benchmarks are the next frontier,›
    LLM score 82 · 7 months ago
  7. Alex Zhang / MIT CSAIL PhD
    2407.
    We need more super hard tasks to properly eval our models!›
    LLM score 92 · 7 months ago
  8. Sherwin Wu / OpenAI API, Head of Engineering
    2408.
    Of all the launches in Codex today, this one was the most game-changing for me – computer use, but without taking over your whole screen!›
    LLM score 92 · 7 months ago
  9. Hyung Won Chung / Research Scientist at Meta
    2409.
    Health care was built for a world where humans were the only interpreters of health data.›
    LLM score 92 · 7 months ago
  10. Chelsea Finn / Physical Intelligence Cofounder
    2410.
    LLM post-training used to mean fine-tuning to a downstream task›
    LLM score 92 · 7 months ago
  11. Karol Hausman / Physical Intelligence Cofounder
    2411.
    Today we're releasing our newest model π0.7: https://t.co/lk5H3Yv7Ru›
    LLM score 92 · 7 months ago
  12. Boris Cherny / Creator of Claude Code
    2412.
    Dogfooding Opus 4.7 the last few weeks, I've been feeling incredibly productive.›
    LLM score 92 · 7 months ago
  13. Ben Burtenshaw / Hugging Face Researcher
    2413.
    crucial new benchmark dropped for long reasoning we need this because even frontier models still can't think long even with 1M context.›
    LLM score 92 · 7 months ago
  14. Tri Dao / Chief Scientist at Together
    2414.
    The dynamical system view gives very clean conditions for looped transformer to be stable https://t.co/md8fispK0Q
    LLM score 92 · 7 months ago
  15. Merve Noyan / Hugging Face ML Engineer
    2415.
    new open-source Bonsai models are out 🔥 > ternary weights in 8B (1.75 GB), 4B (0.86 GB), and 1.7B (0.37 GB)›
    LLM score 25 · 7 months ago
  16. Sergey Levine / Physical Intelligence Cofounder
    2416.
    We finished evaluating π0.7, our new model at Physical Intelligence.›
    LLM score 95 · 7 months ago
  17. Andrew White / Edison Scientific Cofounder
    2417.
    Been using Opus 4.7 for a bit on vulnerability testing (mostly because of the unrelated mythos hype).›
    LLM score 92 · 7 months ago
  18. Jason Wei / AI Researcher at Meta
    2418.
    Beautifully written piece by @FAbnousi about how AI for health might look like in the future›
    LLM score 85 · 7 months ago
  19. Boris Cherny / Creator of Claude Code
    2419.
    Opus 4.7 feels more intelligent, agentic, and precise than 4.6.›
    LLM score 92 · 7 months ago
  20. Sander Dieleman / DeepMind Research Scientist
    2420.
    A while back we had the "rotation trick" to improve VQ bottlenecks (https://t.co/w4XVeN9L0J), now we have DiVeQ, which seems to improve codebook coverage quite significantly.›
    LLM score 92 · 7 months ago
  21. Sayak Paul / Hugging Face Researcher
    2421.
    Working at Hugging Face over the past 3.5+ years has allowed me to identify what technical areas truly interest me!›
    LLM score 65 · 7 months ago
  22. Boris Cherny / Creator of Claude Code
    2422.
    Opus 4.7 is in Claude Code today.›
    LLM score 92 · 7 months ago
  23. Ben Burtenshaw / Hugging Face Researcher
    2423.
    port any transformers model to mlx https://t.co/FtS6gnpVID
    LLM score 92 · 7 months ago
  24. Sander Dieleman / DeepMind Research Scientist
    2424.
    FlexTok/Semanticist provided an elegant recipe to learn semantically coarse-to-fine sequence representations of images.›
    LLM score 92 · 7 months ago
  25. Damek Davis / Assoc. Professor Wharton Stats
    2425.
    I wanted to understand the proof and ended up making a video using opus 4.6.›
    LLM score 92 · 7 months ago
  26. Sander Dieleman / DeepMind Research Scientist
    2426.
    This work provides theoretical grounding for some of the design decisions (cross-entropy loss, learnable embeddings, self-conditioning, entropy-based schedule) in CDCD (https://t.co/B6tYr2AOUk), and brings it into the modern era.›
    LLM score 92 · 7 months ago
  27. Lewis Tunstall / Hugging Face Researcher
    2427.
    New video from the HF journal club: we discuss Apple's self-distillation paper 🍎›
    LLM score 92 · 7 months ago
  28. Merve Noyan / Hugging Face ML Engineer
    2428.
    this is the year of 3D World models 🔥 > Lyra 2.0 by NVIDIA: image → 3D world with Gaussians, 14B params, built on WAN-14B ›
    LLM score 85 · 7 months ago
  29. Michael Elabd / DeepMind Researcher
    2429.
    Continual learning for MLE just hit a new milestone 🥁›
    LLM score 92 · 7 months ago
  30. Cameron Wolfe / Researcher at Netflix
    2430.
    Strongly recommend the LLM-as-a-Verifier writeup.›
    LLM score 92 · 7 months ago
More