CODE - AI Dynamics

Mistral dropped Devstral, a new coding model built with All Hands AI —Small enough to run Macs or a single GPU —Outperforms larger rivals in software engineering —Tuned for agentic software dev
—Available under Apache 2.0 license

→ View original post on X — @rowancheung

22 May 2025

Gemini 2.5 Pro achieves state-of-the-art across math coding multimodality

By

@lmthang

–

22 May 2025 6h32

Not only math, Gemini 2.5 Pro with #DeepThink also achieved state-of-the-art performances for coding (on LiveCodeBench v6), multimodality (on MMMU), and much more! Check out all other #GoogleIO amazing developments with Gemini 2.5 from @GoogleDeepMind and @Google at

→ View original post on X — @lmthang

22 May 2025

Deep Think Model Advances Multi-Domain AI Reasoning Capabilities

By

@oriolvinyalsml

–

22 May 2025 3h18

Deep Think uses new techniques, allowing the model to consider multiple hypotheses before responding. It's not just about math; it also gets an impressive score on LiveCodeBench , MMMU & more! Learn more:

→ View original post on X — @oriolvinyalsml

22 May 2025

Integer Quantification Innovation: Technical Complexity vs Accuracy Trade-offs

By

@fchollet

–

22 May 2025 3h10

The creator reinvents integer quantification from scratch, which is pretty cool! To note, if the goal is just to get 92% accuracy then the hidden layer can be removed (92% is achievable with just logistic regression). But then the project is less technically interesting. IMO ~16

→ View original post on X — @fchollet

22 May 2025