Making LLMs run efficiently can feel scary, but scaling isn’t magic, it’s math! We wanted to demystify the “systems view” of LLMs and wrote a little textbook called “How To Scale Your Model” which we’re releasing today. 1/n

New paper: Reasoning models like DeepSeek  R1 surpass typical LLMs on many tasks. Do they also provide more faithful explanations? Testing on a benchmark, we find reasoning models are much more faithful. It seems this isn't due to specialized training but arises from RL🧵

Today I’m launching Reflection AI with my friend and co-founder Ioannis Antonoglou. Our team pioneered major advances in RL and LLMs, including AlphaGo and Gemini. At Reflection, we're building superintelligent autonomous systems. Starting with autonomous coding.

📣📣 My team at Google DeepMind is hiring a student researcher for summer/fall 2025 in Seattle! If you're a PhD student interested in getting deep RL to (finally) work reliably in interesting domains, apply at the link below and reach out to me via email so I know you aplied👇

Our team in London is hiring! If you want to come work with a wonderful group of researchers on investigating the frontiers of autonomous open-ended agents that help humans be better at doing things we love, come have a look. Link in post below 👇

At Reliant we've found RL to be incredibly efficient at improving answer quality to life sciences' hardest questions. Today we're putting out our work on LLM fine-tuning with off-policy RL, matching llama 70B performance with an 8B model - take a look! arxiv.org/abs/2503.14286

New work out now! 🚀 "The Hardness of Validating Observational Studies with Experimental Data" explores fundamental limits of combining experimental & observational data, and introduces a novel Gaussian Process approach. #AISTATS2025. #CausalInference #ML arxiv.org/abs/2503.14795

We are launching Gemini-pro-exp-03-25 which is our most capable model and generally useful on a wide vareity of real-world tasks. It's #1 on LMArena and SOTA on a wide set of benchmarks. This is a massive Gemini wide effort and I am incredibly proud of the team behind this. 🚀🚀

We have a senior postdoc position available with OATML_Oxford (closing 19/05) to lead work on LLM based causal reasoning with GSK. Please share with anyone you think this might be relevant to! my.corehr.com/pls/uoxrecruit…

Αν προσλαμβάνετε και είστε cool, βάλτε την αγγελία σας στο coolestjob τελεία τζι αρ!

Ethan Dyer and I have started a new team at Anthropic — and we’re hiring! Our team is organized around the north star goal of building an AI scientist: a system capable of solving the long-term reasoning challenges and core capabilities needed to push the scientific

Giving your models more time to think before prediction, like via smart decoding, chain-of-thoughts reasoning, latent thoughts, etc, turns out to be quite effective for unblocking the next level of intelligence. New post is here :) “Why we think”: lilianweng.github.io/posts/2025-05-…

We’ve developed Gemini Diffusion: our state-of-the-art text diffusion model. Instead of predicting text directly, it learns to generate outputs by refining noise, step-by-step. This helps it excel at coding and math, where it can iterate over solutions quickly. #GoogleIO

Introducing our latest update to Gemini 2.5 Pro (06-05), which we expect to become our long term stable release. At a glance: - SOTA on HLE, Aider, and GPQA - Now supports thinking budgets - Same cost, on pareto frontier - Closes gap on 03-25 regressions

We're looking for people to join us to work on Gemini Diffusion and help revolutionize language modeling! Details below: job-boards.greenhouse.io/deepmind/jobs/…

An advanced version of Gemini with Deep Think has officially achieved gold medal-level performance at the International Mathematical Olympiad. 🥇 It solved 5️⃣ out of 6️⃣ exceptionally difficult problems, involving algebra, combinatorics, geometry and number theory. Here’s how 🧵

📜 Today, we’re publishing our latest work in Nature introducing Aeneas, the first AI model for connecting the past.

Gemini 2.5 Deep Think is rolling out to Google AI Ultra subscribers! It uses advances in parallel reasoning & RL to tackle tough math & science problems. Awesome feeling to have IMO medal capabilities at your fingertips! Such a good model!🥇 blog.google/products/gemin…

We’re hiring — βοήθησε μας να πάμε το OpenCouncil από τους 5 στους 50 δήμους. Αν σε ενδιαφέρει ο ρόλος, ή ξέρεις κάποιον που ενδιαφέρεται let me know.

strange object spotted under the microscope over the weekend in the lab...