If anyone needs a video guide to Karpathy's nanochat, check out Stanford's CS336! It covers: - Tokenization - Resource Accounting - Pretraining - Finetuning (SFT/RLHF) - Overview of Key Architectures - Working with GPUs - Kernels and Tritons - Parallelism - Scaling Laws -…
🚨🚨New Paper: Training LLMs to Discover Abstractions for Solving Reasoning Problems Introducing RLAD, a two-player RL framework for LLMs to discover 'reasoning abstractions'—natural language hints that encode procedural knowledge for structured exploration in reasoning.🧵⬇️
Finally had a chance to listen through this pod with Sutton, which was interesting and amusing. As background, Sutton's "The Bitter Lesson" has become a bit of biblical text in frontier LLM circles. Researchers routinely talk about and ask whether this or that approach or idea…
.@RichardSSutton, father of reinforcement learning, doesn’t think LLMs are bitter-lesson-pilled. My steel man of Richard’s position: we need some new architecture to enable continual (on-the-job) learning. And if we have continual learning, we don't need a special training…
29 Oct 2025 အဘိုး တော်တဲ့ ဘုန်းကြီး ဦးဉေယျာဘိဝံသ ပျံတော်မူ။ ဖခင်ဖက်က ရပ်ဆွေ ရပ်မျိုး သဘော၊ အတိအကျ တော်စပ်ပုံ မသိရ။ သက် ၈၀ ဝါ ၆၀ အဂ္ဂမဟာပဏ္ဍိတ၊ စာတတ်လွန်းလို့ အားကျရ။ သက်ရှည် ဉာဏ်ကြီးလွန်းလို့ သတင်းကို မယုံ၊ တုန်လှုပ်။ မိန့်ခဲ့သလို ဘာလုပ်လုပ် စိတ်ထားတတ်ဖို့က အဓိက။ #RandomThoughts
ကျန်းမာရေးနဲ့ ညီညွတ်အောင် မနေတာ ဝန်ခံရမယ်။ အသား တစ်မျိုးသာ စားရာက အသီးအရွက်၊ အသားငါး၊ ကစီဓါတ် မျှတအောင် စားဖို့ ကြိုးစားနေတယ်။ ဒါတောင် ဆီကြော်တွေ ဖြစ်နေတုန်းပဲ။ #health
LeCun is right, when enrolling into PhD program don’t work on what is a hype topic of today. It was true in 2015 for reinforcement learning, it is true in 2025 for LLMs. The topic of tomorrow won’t be the hype topic of today, find a promising niche tech and work on it instead.…
It's application season, and I'm sharing some of my past application materials: - Academic job market (written in Dec 2024) - PhD fellowship (written in Apr 2023) - PhD admission (written in Dec 2019) on my website (j-min.io)
ဝမ်းနည်းဖို့တောင် အချိန် မရှိဘူးလား ဆိုပြီး ငိုချင်တယ်။ လူ တစ်ယောက်တည်းကနေ နှစ်ကိုယ်ခွဲလို့ ရရင်တော့ ကိုယ့်ဖာသာ နှစ်သိမ့်ဖြစ်မှာ အသေအချာပါပဲ။ #randomthought
Fei-Fei Li (@drfeifei) on limitations of LLMs. "There's no language out there in nature. You don't go out in nature and there's words written in the sky for you.. There is a 3D world that follows laws of physics." Language is purely generated signal.
Today Thinking Machines Lab is launching our research blog, Connectionism. Our first blog post is “Defeating Nondeterminism in LLM Inference” We believe that science is better when shared. Connectionism will cover topics as varied as our research is: from kernel numerics to…
She fought for democracy — now she fights for her life. The junta blocks medical access to Daw Aung San Suu Kyi. WHAT’S HAPPENING IN MYANMAR #WhatsHappeningInMyanmar #FreeAungSanSuuKyi #LetOurLeaderAungSanSuuKyiHeal
The myOCR experiment dataset has also been uploaded to GitHub: github.com/ye-kyaw-thu/my… LU Lab Haggling Face မှာတော့ တင်ထားတာ ကြာပါပြီ။ huggingface.co/datasets/LULab…
I haven’t got good sleep for a while. Quite busy these days. So I spent past two days relaxing after school, had a good sleep and now I am ready again to finish my jobs done. 💪
We published our preprint paper that is the first application of Kolmogorov Arnold Convolution for Text classification. Arxiv Link: arxiv.org/abs/2507.06753
Happiness is highly linked to how focused you are on what you're doing. A wandering mind is an unhappy mind.
NLP is not just LLMs. NLP is not just the number of parameters in a language models. Students appreciated a combination of math, code, applications AND... linguistic examples. NLP is about language processing - and language should not be considered merely as data.
Congratulations to @Yoshua_Bengio on launching @LawZero_ — a research effort to advance safe-by-design AI, especially as frontier systems begin to exhibit signs of self-preservation and deceptive behaviour.
နတ် ဆိုတဲ့ Hypothetical သက်ရှိတွေမှာ ရှိတဲ့ ဒိဗ္ဗစက္ကုမျိုး။ ဖြစ်တန်ရာ လမ်းကြောင်းတွေ ကို လူတွေထက် ပိုပြီး များများ နဲ့ မှန်မှန် ရှာခွင့် ရတာ သူတိုက ပိုပြီး Computing power မြင့်လို့ပါ။ သည် World Model တွေက လူတွေထက် Knowledge သာ တောင် လူသား ဦးနှောက်ရဲ့ Computing power မီပါ့မလား?
World models are next-gen AI systems that "imagine" the future by simulating the world around them. No trial-and-error needed — just learned experience. As @ylecun said, they’re crucial to achieving human-level AI, though it may take a decade to fully unlock their power. ▪️ A…
We released the 𝐦𝐲𝐍𝐄𝐑 (𝟕-𝐭𝐚𝐠𝐬) 𝐜𝐨𝐫𝐩𝐮𝐬, including experimental Jupyter Notebooks and a preprint paper. To our knowledge, this is the first publicly available NER corpus for the Myanmar language. github.com/ye-kyaw-thu/my…
United States 趨勢
- 1. Jayden Daniels 20.5K posts
- 2. Dan Quinn 6,321 posts
- 3. Seahawks 35.5K posts
- 4. Sam Darnold 14.3K posts
- 5. Commanders 47.8K posts
- 6. Jake LaRavia 3,609 posts
- 7. #RaiseHail 8,439 posts
- 8. Bronny 12.3K posts
- 9. Joe Whitt 2,155 posts
- 10. jungkook 588K posts
- 11. Marcus Smart 2,903 posts
- 12. #RHOP 6,427 posts
- 13. Jovic 1,035 posts
- 14. Jaxson Hayes 2,751 posts
- 15. #BaddiesAfricaReunion 5,015 posts
- 16. 60 Minutes 64.6K posts
- 17. Larson 20K posts
- 18. Ware 4,992 posts
- 19. Lattimore 2,473 posts
- 20. Denny 19.4K posts
Something went wrong.
Something went wrong.