AI300
  1. Vahid Kazemi / ex MTS at xAI
    511.
    Perhaps not many understood the significance of the M2.5 release by @MiniMax_AI.›
    LLM score 92 · about 1 month ago
  2. Yingru Li / xAI Researcher
    512.
    remind me about some early works on learnable random perturbation in deep RL and RL theory.›
    LLM score 92 · about 1 month ago
  3. Jerry Tworek / ex OpenAI VP of RL
    513.
    I continue to think that data efficiency is not the right perspective to focus on when thinking about deficiencies of todays models›
    LLM score 92 · about 1 month ago
  4. Asher Spector / Cofounder of Flapping Airplanes
    514.
    strong agree; "it's important to have taste" has always felt like a tautology.›
    LLM score 75 · about 1 month ago
  5. Jeff Dean / Chief Scientist at DeepMind
    515.
    Many have really enjoyed our recent Nano Banana (+Pro) generative models for images and Veo models for video.›
    LLM score 75 · about 1 month ago
  6. Andrew White / Edison Scientific Cofounder
    516.
    After surprisingly long amount of work, our literature agent can finally read figures and tables from >150M papers and patents.›
    LLM score 95 · about 1 month ago
  7. Sander Dieleman / DeepMind Research Scientist
    517.
    Several new methods to shape the latent distributions of autoencoders have popped up recently.›
    LLM score 95 · about 1 month ago
  8. Andrew Lampinen / Research Scientist at DeepMind
    518.
    What is the relationship between memorization and generalization in AI? Is there a fundamental tradeoff? In a new blog post I’ve reviewed some of the evolving perspectives on memorization & generalization in machine learning, from classic perspectives through LLMs.›
    LLM score 92 · about 1 month ago
  9. alphaXiv
    519.
    GLM-5: A new SoTA open weights, that BEATS Gemini-3-Pro!?›
    LLM score 85 · about 1 month ago
  10. Sander Dieleman / DeepMind Research Scientist
    520.
    Neat idea: jointly diffuse pixels and DINO features with separate noise levels.›
    LLM score 92 · about 1 month ago
  11. Merve Noyan / Hugging Face ML Engineer
    521.
    you can just speak to language models with a microphone (natively, no TTS layer, runs on-device)›
    LLM score 85 · about 1 month ago
  12. Jie Tang / Z.ai Cofounder
    522.
    We just uploaded our GLM-5's tech report onto arxiv.›
    LLM score 92 · about 1 month ago
  13. Alex Zhang / MIT CSAIL PhD
    523.
    This is very exciting to see!! I'm hoping they also eventually move sub-agents to be integrated into their code execution sandbox as functions, and finally I will no longer have to compare CC to RLMs :) https://t.co/3hL2pXjqVL
    LLM score 72 · about 1 month ago
  14. Sergey Levine / Physical Intelligence Cofounder
    524.
    Check out Noriaki's thread about a way to get VLAs to run in real time with a fast "edge adapter"! https://t.co/HUzoSsrXzn
    LLM score 75 · about 1 month ago
  15. Fei-Fei Li
    525.
    Order matters in diffusion. Check out our latest work! https://t.co/EL1CQM2xHM
    LLM score 92 · about 1 month ago
  16. Adams Wei Yu / DeepMind Research Scientist
    526.
    It is consistent with what showed last week on codeforces ELO.›
    LLM score 75 · about 1 month ago
  17. Damek Davis / Assoc. Professor Wharton Stats
    527.
    Yes, compared to tools available in April, codex is very good at lean.›
    LLM score 82 · about 1 month ago
  18. Leandro von Werra / Hugging Face Head of Research
    528.
    Asked the web instead of the code agent to make a plot of some data and it came up with a creative use of the website screenshot tool: https://t.co/AhOnpTwBQp
    LLM score 82 · about 1 month ago
  19. Chelsea Finn / Physical Intelligence Cofounder
    529.
    Video gen models make pretty videos, but lack physical accuracy›
    LLM score 92 · about 1 month ago
  20. Lucas Beyer / Meta Researcher
    530.
    So, when I have GitHub Copilot point out a typo in my PR, and I'm lazy, I click "Commit suggestion" and it makes a commit.›
    LLM score 82 · about 1 month ago
  21. Boris Cherny / Creator of Claude Code
    531.
    Sonnet 4.6 is now live in Claude Code.›
    LLM score 85 · about 1 month ago
  22. Boris Cherny / Creator of Claude Code
    532.
    Loving Sonnet 4.6 in Cowork -- great balance of capability, speed, and token efficiency https://t.co/ZQErYmBgtE
    LLM score 65 · about 1 month ago
  23. Lucas Beyer / Meta Researcher
    533.
    I usually look at which benches the small model surpasses its previous big brother, if it's only few, i think it gives a hint as to what they focus on.›
    LLM score 82 · about 1 month ago
  24. Eric Jang / ex VP of AI at 1X Robotics
    534.
    Incredibly impressive. We are in the early days of seeing how much there is yet to be unlocked by the humanoid form factor https://t.co/4dXbAPD48x
    LLM score 25 · about 1 month ago
  25. Lewis Tunstall / Hugging Face Researcher
    535.
    Haha, I tried the car wash riddle on QED-Nano and not only did it get it right, but it made the creative suggestion to avoid driving altogether!›
    LLM score 82 · about 1 month ago
  26. John Carmack
    536.
    The glory work of GPU scheduling is in the frontier data centers with hundreds of thousands of GPUs, but a lot of research work is done with single GPU jobs on modest clusters, and the scheduling leaves much to be desired.›
    LLM score 85 · about 1 month ago
  27. alphaXiv
    537.
    Making LLMs truly learn from its experience "Experiential Reinforcement Learning (ERL)"›
    LLM score 92 · about 1 month ago
  28. Cameron Wolfe / Researcher at Netflix
    538.
    Rubric-based RL is one of the most active topics in AI research because it extends the benefits of large-scale RL training to non-verifiable domains.›
    LLM score 95 · about 1 month ago
  29. Ben Burtenshaw / Hugging Face Researcher
    539.
    bold small model release from @Cohere_Labs to add region focused versions of their Tiny Aya model.›
    LLM score 85 · about 1 month ago
  30. Wenting Zhao / Alibaba Qwen Researcher
    540.
    For Qwen-3.5, We took the training recipe from Qwen3-Coder-Next and scale up model parameters, and this results in a much stronger coding agent.›
    LLM score 92 · about 1 month ago
More