Georgios Bakirtzis

I am an assistant professor at Télécom Paris, Institut Polytechnique de Paris, working on autonomous decision making. My research aims to make progress on fundamental questions arising at the intersection of optimization, control, and reinforcement learning.

I organize the Séminaire Systèmes Complexes (w/ Jérémy Dubut).

Updates

07/2026: Pau and I will visit Inria Lille.
06/2026: Pau will visit collaborators at UC Berkeley, UT Austin, and MIT.
05/2026: I organized the NEXUS@AAMAS workshop.
04/2026: We will host David Fridovich-Keil for 5 months across 2026–28 as part of a Hi! PARIS visiting chair.

Selected work

Controllability in preference-conditioned multi-objective reinforcement learning
P. de las Heras Molins, B. Yalcinkaya, L. Peters, D. Fridovich-Keil, & G. Bakirtzis
NeuS; 2026

Breaking exponential complexity in games of ordered preference: A tractable reformulation; D. H. Lee, J. Li, L. Peters, G. Bakirtzis, & D. Fridovich-Keil
preprint; 2026

Categorical semantics of compositional reinforcement learning
G. Bakirtzis, M. Savvas, & U. Topcu
JMLR; 2025

Students

Visitors

David Fridovich-Keil
Hi! PARIS visiting chair

Alumni

Eric Roy-Almonacid   MSc 2026; now at ASDT

Essays

Responsible development for the physical world   Increment; 2021
a case for designing software with safety and formal methods in mind

Turing categories   The n-Category Café; 2019
what is the connection between computation and category theory?

More about me

Contact me  |  Hack my code  |  Read my preprints and publications