AI300
  1. Ben Burtenshaw / Hugging Face Researcher
    31.
    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 · 4 days ago
  2. Ronak Malde / DeepMind Researcher
    32.
    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 · 4 days ago
  3. Cameron Wolfe / Researcher at Netflix
    33.
    Researcher who studied continual learning in grad school revisits literature and its relevance to LLMs amid AGI hype.
    LLM score 70 · 4 days ago
  4. Geoffrey Hinton
    34.
    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 · 5 days ago
  5. Ben Burtenshaw / Hugging Face Researcher
    35.
    The only thing I don't get is why it has to be a Mac mini. Like, is it a physical, sense of power thing? What's wrong with a container, VM, cloud instance, HF space?
    LLM score 20 · 5 days ago
  6. Shane Legg / Chief AGI Scientist at DeepMind
    36.
    @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 · 6 days ago
  7. Boris Cherny / Creator of Claude Code
    37.
    @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
  8. Michael Elabd / DeepMind Researcher
    38.
    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
  9. Andrew Lampinen / Research Scientist at DeepMind
    39.
    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 · 7 days ago
  10. Yoshua Bengio
    40.
    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 · 7 days ago
  11. Shane Legg / Chief AGI Scientist at DeepMind
    41.
    @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 · 7 days ago
  12. Shane Legg / Chief AGI Scientist at DeepMind
    42.
    @GDalegend @korbencopy We get to 2028 and have something quite close to Minimal AGI, but not quite.
    LLM score 20 · 7 days ago
  13. Shane Legg / Chief AGI Scientist at DeepMind
    43.
    @tablefourthree @BogdanIonutCir2 @korbencopy It's actually surprising to me that my probability hasn't shifted yet, given that 2028 is so close. I would have thought that by now...
    LLM score 20 · 7 days ago
  14. Shane Legg / Chief AGI Scientist at DeepMind
    44.
    @korbencopy 50% chance of Minimal AGI by 2028. As I've been saying publicly since 2009.
    LLM score 70 · 7 days ago
  15. Ronak Malde / DeepMind Researcher
    45.
    DeepMind maxed SWE-bench, but lacks tough benchmarks matching real AI use; APEX emerges as the new frontier for frontier models.
    LLM score 75 · 8 days ago
  16. Noam Brown / OpenAI Research Scientist
    46.
    @_sholtodouglas @jekbradbury @GoogleDeepMind @andy_l_jones @OpenAI @kevin_wang3290 A word on comp: I know folks who become quants to make money but 5 years later ask what they're...
    LLM score 20 · 8 days ago
  17. Noam Brown / OpenAI Research Scientist
    47.
    @_sholtodouglas @jekbradbury @GoogleDeepMind @andy_l_jones Labs like @OpenAI also hire researchers straight out of undergrad, like @kevin_wang3290, though the bar is high. Kevin...
    LLM score 60 · 8 days ago
  18. Noam Brown / OpenAI Research Scientist
    48.
    @_sholtodouglas @jekbradbury @GoogleDeepMind @andy_l_jones was a semi-retired quant when he wrote a paper comparing the impact of scaling pretraining vs scaling test-time compute...
    LLM score 70 · 8 days ago
  19. Demis Hassabis / CEO of DeepMind
    49.
    We’re excited to be working with @JNJInnovation to accelerate the path to new medicines. This collaboration brings @IsomorphicLabs' AI drug design engine together with J&J’s...
    LLM score 60 · 9 days ago
  20. Igor Babuschkin / Cofounder of xAI
    50.
    @MillionInt The engineer presented a single grain of etched silicon. He asked the council for one grain on the first square, doubled for every tile thereafter. The council agreed,...
    LLM score 65 · 10 days ago
  21. Keller Jordan / OpenAI Researcher
    51.
    @supreme__sharma There's no good answer. It is brutally difficult for an individual. Either replicate everything yourself or try to find networks outside of X that have a higher...
    LLM score 20 · 11 days ago
  22. Keller Jordan / OpenAI Researcher
    52.
    @eliebakouch In terms of hypothetical disagreement-winrates between these different cultures, I’d estimate that my (heavily muted) FYP has +500 ELO vs. random arXiv papers, and...
    LLM score 30 · 11 days ago
  23. Keller Jordan / OpenAI Researcher
    53.
    The internet has become real life only in the sense that *public* cultural systems have been largely replaced by it Whereas for the activity of making GPUs, the private NVDA...
    LLM score 20 · 11 days ago
  24. Boris Cherny / Creator of Claude Code
    54.
    @aref_vc The formula for the auto-compact buffer is really simple: (contextWindow - maxOutputTokens) - (small constant) By default, the buffer is 45k. If it’s larger than that...
    LLM score 80 · 12 days ago
  25. Boris Cherny / Creator of Claude Code
    55.
    @Wind_AndSnow Ah, you want permissions: { allow: "Read(~/.config/runmenu/*)"}
    LLM score 80 · 12 days ago
  26. Boris Cherny / Creator of Claude Code
    56.
    @aref_vc Reduce your CLAUDE_CODE_MAX_OUTPUT_TOKENS or rm the env var, and you'll decrease the buffer to 45k
    LLM score 70 · 12 days ago
  27. Boris Cherny / Creator of Claude Code
    57.
    @DominikPeters Claude usually only uses Plan agents. The subagents generally result in better plans, though it's a thing we're actively tinkering with
    LLM score 70 · 12 days ago
  28. Boris Cherny / Creator of Claude Code
    58.
    Now in Claude Code: when you accept a plan, Claude automatically clears your context, so your plan gets a fresh context window. We found this helps keep Claude on track longer,...
    LLM score 70 · 12 days ago
  29. Michael Elabd / DeepMind Researcher
    59.
    DeepMind paper flips continual learning on its head: architecture choices fight forgetting better than fixed backbones and algorithms.
    LLM score 85 · 13 days ago
  30. Demis Hassabis / CEO of DeepMind
    60.
    TranslateGemma is our new collection of open translation models for edge devices built on our amazing Gemma 3 open model. They outperform models twice their size at translation...
    LLM score 80 · 14 days ago
More