Florian Hübler

About

I am a doctoral student in computer science at ETH Zurich, working with Niao He in the Optimization and Decision Intelligence Group. My research focuses on optimisation for machine learning, from practical optimiser design to theoretical guarantees. Previously, I completed an MSc in Data Science at ETH Zurich and a BSc in Mathematics at the Technical University of Munich.

I am looking for an industry internship in summer 2027.

Research

I work on optimisation for machine learning, combining the design of practical optimisers with convergence theory and complexity lower bounds. My current focus is understanding and designing matrix-aware optimisers such as Muon and Muown for pre-training large language models. I am also interested in optimisation under heavy-tailed noise and parameter-agnostic (parameter-free) methods for non-convex problems.

Publications and Preprints (Google Scholar)

1. Muown Implicitly Performs Angular Step-size Decay
   F. Hübler^*, K. Lion^*, A. Orvieto, N. He
   Preprint, 2026. arXiv · code
2. Muown: Row-Norm Control for Muon Optimization
   K. Lion, F. Hübler, B. Li, A. Orvieto, N. He
   Preprint, 2026. arXiv · code
3. Free Heavy-Tailed Lunch for Muon: A Theoretical Justification of Empirical Success
   F. Hübler, T. Pethick, S. Sra
   NeurIPS 2026 (Spotlight, top 1.3% of submissions). arXiv · code
4. Can SGD Handle Heavy-Tailed Noise?
   I. Fatkhullin, F. Hübler, G. Lan
   SIAM Journal on Optimization, to appear. OPT 2025 workshop at NeurIPS (Oral). arXiv
5. From Gradient Clipping to Normalization for Heavy Tailed SGD
   F. Hübler^*, I. Fatkhullin^*, N. He
   AISTATS 2025. PMLR 258:2413–2421. PMLR
6. Parameter-Agnostic Optimization under Relaxed Smoothness
   F. Hübler, J. Yang, X. Li, N. He
   AISTATS 2024. PMLR 238:4861–4869. PMLR · code

^*Equal contribution.

Talks

• August 2026. Swiss Optimization Symposium, Monte Verità, Ascona. Contributed talk.
• July 2026. EUROPT Conference on Advances in Continuous Optimization, Linz. Invited talk.
• June 2026. SIAM Conference on Optimization, Edinburgh. Contributed talk.
• February 2026. Rising Stars in AI Symposium, KAUST, Thuwal. Selected talk.
• December 2025. OPT workshop at NeurIPS (Oral), San Diego.

Teaching

• Autumn 2026. Optimization for Data Science, ETH Zurich.
• Spring 2026. AI for Mathematics and Optimization, ETH Zurich.
• Spring 2025. Big Data for Engineers, ETH Zurich.
• Spring 2022. Convex Optimization, ETH Zurich.
• Winter 2020/21. Algebra 1, Technical University of Munich.
• Winter 2019/20. Fundamentals of Programming, Technical University of Munich.

Supervision

• Spring 2026. Bachelor's thesis, Simon Ganter. Optimal Allocation in Pipeline Parallel Distributed Optimization.
• Autumn 2025. Semester project, Oliver Pitsch. Formalising Non-Convex Stochastic Gradient Descent in Lean.
• Spring 2025. Bachelor's thesis, Oliver Pitsch. Formalising Non-Convex Gradient Descent in Lean.

Service

• Area chair. AI for Math workshop at ICML 2026.
• Organising team. Swiss Optimization Symposium, Monte Verità, Ascona, 2026.
• Journal reviewer. Mathematical Programming, SIAM Journal on Optimization, IEEE Transactions on Information Theory.
• Conference reviewer. NeurIPS (2026, 2025, 2024), OPT workshop at NeurIPS (2026, 2025). Top Reviewer at NeurIPS in 2024 and 2025.

Contact

Email

florian.huebler (at) inf.ethz.ch

GitHub

github.com/fhueb

Google Scholar

profile

LinkedIn

profile