George Muraru

RE @googledeepmind,
Former teaching Assistant @upb1818
Former member of the SMPC Team @openminedorg,
Opinions are my own

[Demis Hassabis @demishassabis]

Gemini 2.5 Pro is an awesome state-of-the-art model, no.1 on LMArena by a whopping +39 ELO points, with significant improvements across the board in multimodal reasoning, coding & STEM. You can try it out now in AI Studio https://t.co/lLpF8ToTVJ & @GeminiApp with Gemini Advanced

[Google DeepMind @GoogleDeepMind]

Think you know Gemini? 🤔 Think again.

Meet Gemini 2.5: our most intelligent model 💡 The first release is Pro Experimental, which is state-of-the-art across many benchmarks - meaning it can handle complex problems and give more accurate responses.

Try it now → https://t.co/sxHWaCUzRp

[Arena.ai @arena]

BREAKING: Gemini 2.5 Pro is now #1 on the Arena leaderboard - the largest score jump ever (+40 pts vs Grok-3/GPT-4.5)! 🏆

Tested under codename "nebula"🌌, Gemini 2.5 Pro ranked #1🥇 across ALL categories and UNIQUELY #1 in Math, Creative Writing, Instruction Following, Longer Query, and Multi-Turn!

Massive congrats to @GoogleDeepMind for this incredible Arena milestone! 🙌

More highlights in thread👇

[Soham De @sohamde_]

Two months back, we released a 9B RecurrentGemma model, one of the strongest SSM-based language models out there, trained on 2T tokens!

I finally updated arXiv with some of our results: https://t.co/OACi24CT7w

Link to weights and code for our models in thread!

[Vaibhav (VB) Srivastav @reach_vb]

Welcome RecurrentGemma 9B 🔥

> Same performance as Gemma with more than 25% lower latency and 6-7x higher tokens/ sec ⚡
> Base (9B) and Instruct (9B-IT) models released.
> MMLU - 60.5, CommonSenseQA 73.2, AGIEval 39.3 - pretty strong base model to fine-tune further.
> Based on the Griffin Architecture
> Achieves faster inference with long sequences by replacing gloabal attention with local and linear recurrences.

> Available in Transformers! 🤗

Massive Kudos to Google for continue open research for alternative architectures! GG!

[Caglar Gulcehre @caglarml]

Yay 🎉! I am really happy that the 2B version of our Griffin architecture is finally open-source! Great work and job the whole team for doing this. I have already started playing with it in colab ☺️

[Aleksandar Botev @botev_mg]

Following our previous work, we are releasing RecurrentGemma - a fully open source 2B model based on our Griffin architecutre!

Code + weights as everyone has wished for!

Code on Github: https://t.co/K4Ifc8e4Oe
Weights on Kaggle: https://t.co/wDntCLaZ8Q

[Soham De @sohamde_]

Releasing RecurrentGemma - one of the strongest 2B-param open models designed for fast inference on long sequences and massive throughput!

Both pre-trained and IT checkpoints available + code - try them out here!

Code: https://t.co/VskJw7eo9Z
Weights: https://t.co/MvdkGhlqDS

[Samuel L Smith @SamuelMLSmith]

Announcing RecurrentGemma!
https://t.co/nR6TGrnWQK

- A 2B model with open weights based on Griffin
- Replaces transformer with mix of gated linear recurrences and local attention
- Competitive with Gemma-2B on downstream evals
- Higher throughput when sampling long sequences

[Soham De @sohamde_]

Just got back from vacation, and super excited to finally release Griffin - a new hybrid LLM mixing RNN layers with Local Attention - scaled up to 14B params!

https://t.co/FDyBXyLzAV

My co-authors have already posted about our amazing results, so here's a 🧵on how we got there!

[Aleksandar Botev @botev_mg]

We present Griffin: A hybrid model mixing a gated linear recurrence with local attention. This combination is extremely effective: it preserves all the efficient benefits of linear RNNs and the expressiveness of transformers. Scaled up to 14B!
https://t.co/uptKapilDM

[AK @_akhaliq]

Human-Timescale Adaptation in an Open-Ended Task Space
abs: https://t.co/9UqYIs6qEE

[PyTorch @PyTorch]

If you installed PyTorch-nightly on Linux between Dec. 25 and Dec. 30, uninstall it and torchtriton immediately and use the latest nightly binaries.

Read the security advisory here: https://t.co/jnCSGXJRY0

[Vitaly 🇺🇦 Feldman @vitalyFM]

🧵Finally, the promised follow up on the privacy paper that got the @icmlconf '22 outstanding paper award. With Nicholas Carlini and @srxzr we had a closer look at the paper and wrote up the summary of the flaws we found. https://t.co/PJ479x6Zti

[OpenMined @openminedorg]

We're very excited to work w/ New Zealand & US Govs, @Microsoft, & @Twitter on expanding #PySyft use helping independent researchers study major #AI algorithms

👇New Zealand PM @jacindaardern & French President @EmmanuelMacron announce the project

WATCH: https://t.co/bmQsISCRWL

[OpenMined @openminedorg]

We’re very excited to announce a new partnership with Twitter to advance algorithmic transparency using PySyft!

https://t.co/3yFAF9PtaF

[George Muraru @GeorgeMuraru]

@ilyaraz2 Til you drop - ItaloBrothers, Paint the wall - Killrude, Free Animal - Foreign Air

[George Muraru @GeorgeMuraru]

@eigenhector My mind went directly to Braess's Paradox

[OpenMined @openminedorg]

NEW POSITION: Secure MPC - Cryptography Engineer

This is a role primarily involved in creating distributed systems which implement SMPC protocols within AI/ML tools!

- 100% remote
- paid/volunteer positions available
- 100% open source codebase

Details: https://t.co/LDbBs22Lg7

[George Muraru @GeorgeMuraru]

@FitFounder Never reply when you are angry, never make a promise when you are happy.