AI300
  1. Ethan Shen / Ai2 Researcher
    1.
    SWE-Bench demands multiple seeds for statistically significant results; single-seed scores, often reported, are meaningless.
    LLM score 95 · 2 days ago
  2. Cameron Wolfe / Researcher at Netflix
    2.
    Trinity Large is a sparse 400B-A13B MoE with 256 experts (4 active per token), chosen for a ~30-day training sprint that also yields blazing speed plu
    LLM score 85 · about 5 hours ago
  3. Ethan Shen / Ai2 Researcher
    3.
    We experiment with mixing SERA's first and second rollout data, finding they complement each other to boost performance in data-constrained regimes de
    LLM score 85 · 2 days ago
  4. Ethan Shen / Ai2 Researcher
    4.
    Best data comes from truncated trajectories that just barely exceed the context limit.
    LLM score 85 · 2 days ago
  5. Ronak Malde / DeepMind Researcher
    5.
    Stanford/Nvidia's TTT+RL paper trains an open-source model that beats AlphaEvolve, sets new Erdos overlap bound, and crafts 2x faster A100 kernels.
    LLM score 85 · 3 days ago
  6. Michael Elabd / DeepMind Researcher
    6.
    Memory is the biggest blocker to continual learning: how do we make models selectively remember, forget mistakes, and surface the right context like h
    LLM score 85 · 6 days ago
  7. Ethan Shen / Ai2 Researcher
    7.
    GA uses 4.5 Air as teacher; SERA-32B uses GLM 4.6. GLM 4.7 Flash flops at 32K/64K context, possibly too reliant on long contexts.
    LLM score 85 · 1 day ago
  8. Ben Burtenshaw / Hugging Face Researcher
    8.
    Claude teaches open models to write CUDA kernels via agentic capability transfer, detailed in this blog post.
    LLM score 85 · about 22 hours ago
  9. Tim Dettmers / Research Scientist at Ai2
    9.
    Our method became so efficient (26x vs RL; 57x vs other synth gen), that we could easily generate 1000s of trajectories for a single repo. This makes the coding agents very...
    LLM score 85 · 2 days ago
  10. alphaXiv
    10.
    Paper shows LLM in sandbox with browsing, file ops, and code execution unlocks general agentic intelligence and boosts non-code performance—sans train
    LLM score 85 · 1 day ago
  11. Andrej Karpathy / AI researcher
    11.
    @ChiragLathiya The nearest neighbor really is some kind of a junior engineer. Its ideas about what experiments to run on e.g. nanochat have been surprisingly bad, but its...
    LLM score 80 · 3 days ago
  12. Tim Dettmers / Research Scientist at Ai2
    12.
    From there we could do a massiv amounts of experiments and really understand what matters for training coding agents. The most important insights came from carefully evaluating...
    LLM score 80 · 2 days ago
  13. Niklas Muennighoff / AI Researcher at Stanford
    13.
    Community-built open benchmarks work really well, e.g., Terminal-Bench, HLE, MMTEB. Anyone can improve them by adding tasks without breaking others. If models had the same...
    LLM score 80 · 2 days ago
  14. Ronak Malde / DeepMind Researcher
    14.
    DeepMind maxed SWE-bench, but lacks tough benchmarks matching real AI use; APEX emerges as the new frontier for frontier models.
    LLM score 75 · 7 days ago
  15. Andrew Lampinen / Research Scientist at DeepMind
    15.
    Is it better to go to academia or industry for research in AI or cognitive science? It's the most frequent question I get asked by PhD students, and I've finally written up some...
    LLM score 75 · 6 days ago
  16. Tim Dettmers / Research Scientist at Ai2
    16.
    This is very impactful: you can now distill frontier performance into small models that are specialized to private repositories. Companies can quickly and cheaply train on their...
    LLM score 75 · 2 days ago
  17. alphaXiv
    17.
    New interpretability idea 'Patterning': reverse-engineer training data that induces desired model circuits/algorithms by measuring internal sensitivit
    LLM score 75 · about 24 hours ago
  18. Shane Legg / Chief AGI Scientist at DeepMind
    18.
    @korbencopy 50% chance of Minimal AGI by 2028. As I've been saying publicly since 2009.
    LLM score 70 · 6 days ago
  19. Shane Legg / Chief AGI Scientist at DeepMind
    19.
    @kevinkshah @korbencopy I'd give it a 50% probability of happening within 5 years of Minimal AGI. I'd be surprised if it's less than 1 or more than 10 years.
    LLM score 70 · 6 days ago
  20. Andrej Karpathy / AI researcher
    20.
    @0xabi96 It feels like I’m cheating. Which is a very weird feeling to have. It takes a while to unpack. It’s because some code that used to be a point of pride and high IQ and...
    LLM score 70 · 3 days ago
  21. Tim Dettmers / Research Scientist at Ai2
    21.
    his work was mostly the genius of Ethan Shen. He started without a background in coding agents and is now one of the most knowledgeable people in the coding agent area. He has all...
    LLM score 70 · 2 days ago
  22. Noam Brown / OpenAI Research Scientist
    22.
    1987: AI can't win at chess—planning is uniquely human 1997: AI can't win at Go—intuition is uniquely human 2016: AI can't win at poker—bluffing is uniquely human 2023: AI...
    LLM score 70 · 3 days ago
  23. Cameron Wolfe / Researcher at Netflix
    23.
    Researcher who studied continual learning in grad school revisits literature and its relevance to LLMs amid AGI hype.
    LLM score 70 · 4 days ago
  24. Cameron Wolfe / Researcher at Netflix
    24.
    Standard RL excels at continual learning, suggesting it's closer for LLMs than thought.
    LLM score 65 · 2 days ago
  25. Andrej Karpathy / AI researcher
    25.
    @airesearch12 💯 @ Spec-driven development It's the limit of imperative -> declarative transition, basically being declarative entirely. Relatedly my mind was recently blown...
    LLM score 60 · 3 days ago
  26. Yoshua Bengio
    26.
    Il est essentiel de développer à la fois une réglementation efficace, des garde-fous sociaux et des solutions techniques comme celles que nous développons à @LawZero_ pour éviter...
    LLM score 40 · 6 days ago
  27. Boris Cherny / Creator of Claude Code
    27.
    @ariccio Use it for when the main agent doesn't need to see the work that a skill did, only its output. eg. I use it for /commit-push-pr, since the main agent doesn't need to see...
    LLM score 40 · 6 days ago
  28. Geoffrey Hinton
    28.
    I just watched a really great conversation about the future of AI. Every politician should watch it before they join the lemmings saying that regulation of AI will interfere with...
    LLM score 30 · 4 days ago
  29. Ben Burtenshaw / Hugging Face Researcher
    29.
    I know it's massively off trend, but I love setting up a nice quiver of models in @cursor_ai, reviewing generated tokens in the legit ui, switching models for certain parts, and...
    LLM score 30 · 3 days ago
  30. Shane Legg / Chief AGI Scientist at DeepMind
    30.
    @2024dion @deredleritt3r The fact that it still seems 50-50 by 2028 surprises me! I would have thought that evidence would have tipped it one way or another by now.
    LLM score 30 · 5 days ago
More