Gonna try to pin a few favorite posts that linger in mind over time:
Amusing how 99% of people using their own brains forget how it works: The brain is an advanced probability machine. It keeps predicting the next most likely thought, word, or action based on incoming signals and past learning. Under the hood, billions of neurons are doing…
Holy shit... Tencent researchers just killed fine-tuning AND reinforcement learning in one shot 😳 They call it Training-Free GRPO (Group Relative Policy Optimization). Instead of updating weights, the model literally learns from 'its own experiences' like an evolving memory…
I'm super excited about M5. It's going to help a lot with compute-bound workloads in MLX. For example: - Much faster prefill. In other words time-to-first-token will go down. - Faster image / video generation - Faster fine-tuning (LoRA or otherwise) - Higher throughput for…
Highly recommend this book if you dare to get into the world of biophysical thinking. Years ago Jorge, now @FIU, told me to get back to school and learn differential equations. I’m so glad he did. Biology needs math, so does medicine.
Finally, someone has solved a real problem with AI! No more having to take a paper in the format for a journal that rejected you, and reformat it for a new journal. Well done!! formatmypaper.com
Meta just did the unthinkable. They figured out how to train AI agents without rewards, human demos, or supervision and it actually works better than both. It’s called 'Early Experience', and it quietly kills the two biggest pain points in agent training: → Human…
Everyone wants to get into robotics. No one knows where to start. LeRobot's Francesco just dropped a 70-page crash course that takes you from zero to cutting-edge: - RL sim/real - ACT, Diffusion policies - VLAs, SmolVLA, Pi-0 Absolute gold if you want to catch up fast.
A comprehensive, hands-on tutorial on the most recent advancements in robotics 🤟 ...with self-contained explanations of modern techniques for end-to-end robot learning & ready-to-use code examples using @LeRobotHF and @huggingface. Now available everywhere! 🤗
glm 4.6 on M3 Ultra (8bit, MLX) is smarter than GPT-5 in some domains I tested it on! 🔥
Not a single YouTuber who received the NVIDIA DGX Spark has compared it to the 14-inch MacBook Pro M4 Max with 40 cores and 128GB of RAM even though they’re priced almost the same. The reason? The MacBook Pro has a bandwidth of 546 GB/s, while the DGX Spark only 273 GB/s.
Transformer vs. Mixture of Experts (MoEs):
You're in an ML Engineer interview at MistralAI. The interviewer asks: "We need an LLM that excels across code, math & creative writing. How do you achieve multi-domain performance?" You: "I'll increase the number of attention heads." Interview over. Here's what you missed:
DGX Spark GPT OSS 120B: 11.65 tok/sec M3 max GPT OSS 120B: 41.71 tok/sec This is very bad. Can't be explained from bad ollama performance and slower memory speed (273GB/s vs 400GB/s) alone.
🚀 SGLang In-Depth Review of the NVIDIA DGX Spark is LIVE! Thanks to @NVIDIA’s early access program, SGLang makes its first ever appearance in a consumer product, the brand-new DGX Spark. The DGX Spark’s 128GB Unified Memory and Blackwell architecture set a new standard for…
A microscopic tardigrade, also known as a water bear, walking across a glass slide. Extremely resilient, it can survive decades without food or water and are the only known animal that can survive in direct exposure in space. [📹 Tobin Sparling]
I don’t want to throw shades at anyone but the HRM architecture is better explained in the TRM paper than in its own
we’re finally diving in the inner working of this itsy bitsy tiny reasoning network this sunday btw if you create a new architecture having a clear visual of said arch is awesome many many many thanks alexia
«Unitree robodogs have no use case outside research»
🧠 A new wiring diagram of the 166,000+ neurons in the brain & nerve cord of a male fruit fly is a key tool in uncovering how the brain enables complex behavior—information that could ultimately help scientists understand what causes different diseases ➡️ hhmi.news/4o3EJnk
United States 趨勢
- 1. Westbrook 8,893 posts
- 2. Gold Glove 5,968 posts
- 3. #wednesdaymotivation 8,555 posts
- 4. Jay Jones 66.2K posts
- 5. Sacramento Kings 4,531 posts
- 6. Rickey 1,550 posts
- 7. Veo 3.1 2,970 posts
- 8. Jared Leto 7,365 posts
- 9. Big Balls 11.5K posts
- 10. Young Republicans 128K posts
- 11. Voting Rights Act 15.9K posts
- 12. Vision Pro 5,001 posts
- 13. The Kings 77.3K posts
- 14. Brodie 3,602 posts
- 15. Summer Walker 3,294 posts
- 16. Lavine N/A
- 17. Bessent 72.1K posts
- 18. Justice Jackson 3,047 posts
- 19. Dansby N/A
- 20. Sabonis N/A
Something went wrong.
Something went wrong.