"It is one of the most remarkable things that in all of the biological sciences there is no clue as to the necessity of death." -- Richard Feynman (from The Pleasure of Finding Things Out)

@jeremyphoward @HetPatel____ you might think this is interesting

RT @jeremyphoward: What if I told you that, without needing any task specific fine tuning, the encoder-only ModernBERT 0.3b beats Qwen 0.5b…

RT @iScienceLuvr: Competitive Programming with Large Reasoning Models New paper from OpenAI highlighting results of their reasoning models…

RT @burkov: As promised last week, here's a full tutorial for training a language model using the reinforcement learning algorithm GRPO, th…

RT @kimmonismus: Visual chain of thought. Looks like we get „thinking“- vision models

RT @polynoamial: I’m excited to see academics pursuing radically different approaches to scaling inference compute. RL on CoT is one way, b…

RT @jonasgeiping: Ok, so I can finally talk about this! We spent the last year (actually a bit longer) training an LLM with recurrent d…

RT @tomgoldsteincs: New open source reasoning model! Huginn-3.5B reasons implicitly in latent space 🧠 Unlike O1 and R1, latent reasoning…

RT @Riazi_Cafe_en: Paul Halmos on doing mathematics: "Don't just read it; fight it! Ask your own questions, look for your own examples, d…

RT @garrytan: I'm proud to announce YC's first AI Startup School— June 16 and 17th in San Francisco. It's a totally free conference bringin…

RT @y0b1byte: With interest in RL skyrocketing, I remind you that this resource by @jachiam0 is still the best introduction to RL you can f…

RT @Dr_Singularity: BIG Cancer breakthrough A personalized vaccine for kidney cancer generated strong immune responses in all trial pat…

@3blue1brown Awesome. At one point Tao was going to write a book about the cosmic distance ladder. Glad the project survives in this form.

@SebastienBubeck Not sure what to make of this though:

RT @emollick: This paper is wild - a Stanford team shows the simplest way to make an open LLM into a reasoning model. They used just 1,000…

RT @Sentdex: The obfuscating money launderers were not likely prepared for the guy who used AI to read 2000 year old ancient scrolls withou…

RT @Geiger_Capital: Elon really has his best cracked engineers sleeping in the Eisenhower Building, trying to put some startup energy into…

RT @konstmish: Learning rate schedulers used to be a big mistery. Now you can just take a guarantee for *convex non-smooth* problems (from…

RT @ArthurB: If you traveled back to 2015 and had to give a recipe for AGI, you'd need to convey just three insights, deceptively simple in…