large language models, reasoning, fine-tuning, test-time computation, reinforcement learning with human feedback, world models
News RSS feed
News: new Marek Kudzbel / Podcaster: Podcast conversation in Bratislava. Direct episode link to be added after publication.
News: new Na jednej vlne: Conversation with Lenka Debnárová in Bratislava. Direct episode link to be added after publication.
News: new STVR: Confirmed television debate on AI developments, benefits, risks and broader implications. Episode link to be added after broadcast.
News: new Startitup.sk: Interview at the Carlton Hotel in Bratislava. Direct article or video link to be added after publication.
News: new Invited speaker at University of Wrocław Faculty Seminar, November 2026–February 2027, Faculty of Mathematics and Computer Science, University of Wrocław, Wrocław, Poland.
News: new Speaking at FIIT STU in Bratislava in November 2026 on LLM reasoning, post-training and alignment, followed by Q&A.
News: new Speaking at DigiTec 2026 in Yerevan, Armenia, on November 21, on sovereign AI and European compute. DigiTec 2026
Bio
Michal is the Founding Researcher at Isara Labs, tenured researcher at Inria, and a lecturer at MVA at ENS Paris-Saclay. In 2026, Michal was elected an ELLIS Fellow in the ELLIS Unit Paris. Michal is primarily interested in designing algorithms that would require as little human supervision as possible. He works on methods and settings that are able to deal with minimal feedback, such as deep reinforcement learning, bandit algorithms, self-supervised learning, or self play. Michal has recently worked on representation learning, world models and deep (reinforcement) learning algorithms that have some theoretical underpinning. In the past he has also worked on sequential algorithms with structured decisions where exploiting the structure leads to provably faster learning. Michal is now working on a new generation of large language models (LLMs), in addition to providing algorithmic solutions for their scalable test-time inference, fine-tuning and alignment. He received his PhD in 2011 from the University of Pittsburgh, before getting a tenure at Inria in 2012 and co-creating Google DeepMind Paris with R. Munos. In 2024, he became a Principal Llama Scientist at Meta, building online reinforcement learning stack and research for Llama 3. In 2025, he joined Isara Labs as a founding researcher.
Selected work
Research threads spanning frontier models, representation learning, bandits, and sparsification.
Nash-MD / NLHF
Preference learning as a two-player game, connecting LLM alignment, human feedback, and provable Nash-equilibrium algorithms.
NLHF · Nash-MD · preference learning
IX / implicit exploration
Implicit exploration replaces explicit exploration mixing with a biased loss estimator, giving sharp guarantees under partial and side-observation feedback.
side observations · data-dependent bounds · contextual bandits
Bandits & structured exploration
A long line on learning faster from structure: graphs, side observations, combinatorial actions, and Bayesian exploration.
Spectral · graph bandits · best-arm ID · semi-bandits · Thompson sampling
SQUEAK & sparsification
Single-pass sparsification of kernel matrices and graphs using ridge leverage scores, with scalable distributed variants.
Kernel approximation · ridge leverage scores · spectral sparsification
Coming up
Contact
Paris, France
40 avenue Halley
59650 Villeneuve d'Ascq, France
+33 3 59 57 78 01
4, avenue des Sciences
91190 Gif-sur-Yvette, France