Aleks Ficek 🧪
@AlekFicek
Followers
379
Following
5K
Media
41
Statuses
424
LLM Research @NVIDIA | Also run a meetup @zurichnlp | Solve intelligence, solve everything else
Zurich, Switzerland
Joined June 2013
Thank you @SchmidhuberAI for speaking in front of a packed room at ZurichAI in the @ETH_AI_Center yesterday! It's the biggest event so far, by far. Thanks everyone for coming; we're sorry for anyone who couldn't didn't get a spot. More and bigger things are planned!
7
11
121
Our team has just achieved gold medal at the International Olympiad in Informatics (IOI) with open-weight models for the first time! 🏅 This marks a pivotal step in bridging the gap between open and closed-source AI, see @mehrzadsamadi's post for more details. 💥
(1/5) 🚀 For the first time, an open-weight model has achieved a gold medal at the International Olympiad in Informatics (IOI). We did it with gpt-oss-120b and our new framework, GenCluster!
0
1
4
Excellent opportunity to close the gap between open and closed source LLMs 💥
Do you want to work on LLM and DLM model post-training with us? @JiantaoJ is hiring!
0
0
0
📣 Announcing the release of OpenReasoning-Nemotron: a suite of reasoning-capable LLMs which have been distilled from the DeepSeek R1 0528 671B model. Trained on a massive, high-quality dataset distilled from the new DeepSeek R1 0528, our new 7B, 14B, and 32B models achieve SOTA
12
122
617
As for the future, I'll be looking to get more involved in the AI scene in London so if you have any leads or just want to connect, please reach out! 😄 Thank you to all of the attendees, speakers and the team running the events. ❤️
0
0
0
I will keep telling people that Zurich has the highest tech talent density in the world and if you are in the city, RSVP to one of our meet-ups!
zurichai.ch
ZurichAI is the largest regularly scheduled machine learning meetup in Switzerland. We're in Zurich and host events for NLP, CV & more with 100+ regular attendees.
1
0
0
It has been incredibly fulfilling to run something so grassroots and see the AI community in Zurich flourish. We've hosted 26 meet-ups and have had thousands of people attend talks given by AI researchers to startup founders 🤯
1
0
1
2.5 years ago I co-founded ZurichAI because I felt the city needed something to connect all of the talented AI researchers. I'm moving to London so a week ago I presented my team's work on "Synthetic Data for Coding and Reasoning in LLMs" as a fitting end to my contribution. 🧵
2
0
7
People tell me that there is no active AI/builder community in Switzerland 🇨🇭 Its improving! E.g. join the @zurichnlp events. Today was a nice talk from @AlekFicek about synthetic data generation for reasoning models!
0
1
12
Come check out our next ZurichNLP meetup, where I'll present on generating synthetic data for reasoning and coding in LLMs! 🚀 👉 ZurichNLP Meetup #17 this Tuesday June 17th at 18:00 https://t.co/drVtQgUPLc
zurichai.ch
Aleksander Ficek (NVIDIA) on synthetic generators & verifiers for coding and Matteo Saponati (ETH Zurich) on the structures of self-attention beyond keys and queries.
2
0
7
ZurichNLP#17 will be in two weeks on Tuesday, June 17th at the @ETH_AI_Center! Our very own @AlekFicek (NVIDIA) will talk about synthetic generators and verifiers for coding and @matteosaponati (ETH Zurich) about structures of self-attention beyond keys and queries.
1
5
10
NVIDIA just open sourced Open Code Reasoning models - 32B, 14B AND 7B - APACHE 2.0 licensed 🔥 > Beats O3 mini & O1 (low) on LiveCodeBench 😍 Backed by OCR dataset the models are 30% token efficient than other equivalent Reasoning models Works with llama.cpp, vLLM,
17
135
737
The largest Math dataset for reasoning yet, 5+ M samples, and the strongest scores yet for multiple model sizes from the Math Reasoning team !
Keeps getting better: Nvidia also dropped OpenMath Nemotron 32B & 14B - secured FIRST prize in AIMO-2 competition 🤯 > beats DeepSeek R1, QwQ and more on AIME, HLE-Math and more So cool to see Nvidia not just releasing model checkpoints, but also the code and the datasets too
0
5
14
ZurichCV#9 is in exactly one week on Tuesday 29th at @ETH_AI_Center. Join Linus Scheibenreif (ETH Zurich) and Pascal Chang (Disney Research). RSVP at zurichai(.)ch!
1
5
17
Cool to see Scoring-Verifiers and OpenCodeReasoning there! 🤩
What's cool about @nvidia is that in addition to models, they release tons of cool datasets! Why are the other big tech not doing that too? https://t.co/uM0sHKZ4yI
0
0
2
Good morning world✨We just dropped the best open-source reasoning model. Enjoy! 🫳 https://t.co/Cs0g8lmF13
0
0
8
Nice to see the traction for our work!
OpenCodeReasoning: Competitive Coding Dataset - 735K Python solutions across 28K unique programming problems - Largest reasoning-based synthetic dataset for code generation - Generated by NVIDIA's R1 model with full reasoning steps - Sourced from 10 competitive coding platforms
0
0
4
Markets down 📉 NVIDIA LLM research 📈 Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Models! Better or on-par accuracy compared to other similar open-sourced transformer models while being up to ✨ 3× faster at inference ✨ https://t.co/s9MvKYBHEc
arxiv.org
As inference-time scaling becomes critical for enhanced reasoning capabilities, it is increasingly becoming important to build models that are efficient to infer. We introduce Nemotron-H, a family...
0
5
23