Updates

🪶 We have released Kolibri 1 a 78B-A3B active open-weight model. Check out the tech report for all details. Really proud to have contributed to this European model!

💼 I joined Aleph Alpha Research as an AI Researcher, working on the evaluation of large language models.

🗣️ I gave an invited talk at the Workshop on Distributed Training in the Era of Large Models at KAUST, in Thuwal, Saudi Arabia.

🗣️ I gave an invited talk at ICCOPT in Los Angeles about my work on AlgoPerf.

🤝 We are organizing the CodeML Workshop at ICML 2025 in Vancouver focusing on open-source software in machine learning!

📝 Our paper “Accelerating Neural Network Training: An Analysis of the AlgoPerf Competition” has been accepted at ICLR 2025! See you in Singapore!

🤝 We are organizing the first AlgoPerf Workshop at Meta HQ in Menlo Park on February 11th & 12th.

📝 Our paper “Kronecker-Factored Approximate Curvature for Modern Neural Network Architectures” received a spotlight presentation at NeurIPS 2023!

📝 We published the first paper of our MLCommons’ Algorithms working group on arXiv titled “Benchmarking Neural Network Training Algorithms”.

🎓 I succesfully defended my Ph.D. thesis with the title “Understanding Deep Learning Optimization via Benchmarking and Debugging”!

📝 Our paper “Cockpit: A Practical Debugging Tool for the Training of Deep Neural Networks” has been accepted at NeurIPS 2021.

📝 Our work “Descending through a Crowded Valley - Benchmarking Deep Learning Optimziers” has been accepted at ICML 2021.

🏛️ I have been elected as a co-chair for the MLCommons working group on Algorithmic Efficiency together with George Dahl.

🎖️ I have been distinguished as a top reviewer for ICML 2020 and received a Top 10% Reviewer award for NeurIPS 2020.

📝 Our paper “DeepOBS: A Deep Learning Optimizer Benchmark Suite” has been accepted at ICLR 2019.