←
Mohammad Saffar257 posts
Mohammad Saffar avatar

Mohammad Saffar

@msaffar3

Research Scientist @googledeepmind, Gemini omni | past: founding researcher @reve and Google brain

◷ Joined July 2016▧ 8 Photos and videos
Melvin Johnson avatar
⟳ Mohammad Saffar retweeted
Melvin Johnson@melvinjohnsonp·Nov 10

Today we’re rolling out Gemini 4 Carbon, our strongest model with frontier performance in cybersecurity capabilities and long-horizon agentic workflows. We’ve poured a lot of love into it, and it’s very good at working through hard problems.

Mohammad Saffar avatar
Mohammad Saffar@msaffar3·Nov 6

Making pixels move more

Big news: Gemini Omni 1.2 Flash by @GoogleAI debuts at #1 in both Text-to-Video and Image-to-Video, and #3 in Video Edit! Gemini Omni 1.2 Flash landed #1 in Text-to-Video with 1541 pts. This is a +27 pt improvement from the previous Gemini Omni 1.1 Flash at #2 with 1514 pts. With this release, Google’s Omni variants take all 3 top slots in Text-to-Video. See the Image-to-Video and Video Edit results in thread. Congrats to the @GoogleAI team on this release!

Post attachment
Mohammad Saffar avatar
Mohammad Saffar@msaffar3·Oct 29

Argon to carbon in a few weeks. This is what happens when research starts to accumulate.

Mohammad Saffar avatar
Mohammad Saffar@msaffar3·Oct 28

trust

Introducing Gemini 4 Carbon, it's a big jump over Argon which we released just 2 weeks ago. Gemini 4 Carbon is SOTA on cyber, long horizon tasks, ranks #1 on Artificial Analysis V5, significantly improves on usability in domains like coding, and is more aligned.

Post attachment
Ruben Villegas avatar
⟳ Mohammad Saffar retweeted
Ruben Villegas@RubenEVillegas·Oct 28

@zacharynado right now

Introducing Gemini 4 Carbon, it's a big jump over Argon which we released just 2 weeks ago. Gemini 4 Carbon is SOTA on cyber, long horizon tasks, ranks #1 on Artificial Analysis V5, significantly improves on usability in domains like coding, and is more aligned.

Post attachment
Mohammad Saffar avatar
Mohammad Saffar@msaffar3·Oct 28

We do treat multi modality very seriously 🚀

This is Genie 4, our most realistic and interactive world model yet. 🌐 Explore dynamic environments in real time to see how the scene responds, sounds, and how the people in it move, react, and speak.

2:27
Mohammad Saffar avatar
Mohammad Saffar@msaffar3·Oct 23

This is full of insights!

Hi all, I have a new result I'm excited to share. It turns out that *all existing implementations* of softmax and cross-entropy (that's JAX, PyTorch, textbooks, etc) have a subtle bug --- two bugs, actually. A simple rewrite fixes both, and eliminates "slingshots" when grokking:

0:07
Mohammad Saffar avatar
Mohammad Saffar@msaffar3·Oct 17

Yes, hardware lottery is not when your research ideas are hardware friendly by accident, you should actively design for it.

The evolution of interleaved SWA and global layers with window size 512 happened due to merging Routing Transformer (github.com/google-research/g…) with scan layers - XLA likes repeated blocks to reduce compilation time. Doing top-k sparse attention proved to be harder on TPUs and most of the successful sparsification has happened on GPUs (deepseek, minimax etc).

Mohammad Saffar avatar
Mohammad Saffar@msaffar3·Oct 14

Argon multi modal is 👑

Congrats to @GoogleAI's Gemini 4 Argon for taking the top spot of our Humanity's Sixth Sense (HSS) benchmark with a score of 55.4%, edging out GPT-6-Astra at 53.6%. 🥇 HSS tests AI’s ability to reason beyond what’s visible, from hidden objects to predicting what happens next and understanding human behavior. We took a closer look at performance across categories, and that’s where things get interesting. 👀🧵

0:06
Mohammad Saffar avatar
Mohammad Saffar@msaffar3·Oct 1

Dont mess with @zacharynado ✨

Bee@thtbee_Oct 1

Zachary has become my favorite google employee in just 2 dayss.. He was underground for almost a year, probably just bookmarking posts and from the moment Gemini 4 Argon launched.. he has been ruthless in calling out every negative post.. I just hope that when we get hands-on access to the model.. all the call outs stay valid!! Great job @zacharynado and good luck!!

Post attachment
Mohammad Saffar avatar
Mohammad Saffar@msaffar3·Oct 1

Long context has been a first class citizen problem from the early days of Gemini and not surprisingly Gemini 4 is amazingly good at it. Not only at ingestion but also with long seq generation. It has been a joy seeing Argon working on long horizon and complex tasks.

Mohammad Saffar avatar
Mohammad Saffar@msaffar3·Sep 30

Gemini 4 is a really amazing model. I have been using it internally for a while, specially for AI research and infra work ✨✨✨✨

Mohammad Saffar avatar
Mohammad Saffar@msaffar3·Sep 30

Bitter lesson pilled🚀

Introducing Gemini 4 Argon, our new frontier model, rolling out to cyber defenders starting today, and more widely as soon as possible. I am really excited by the progress we have made here. Argon is priced at $2 in and $10 out during introductory pricing!

Post attachmentPost attachmentPost attachmentPost attachment