About me

Starting in January 2025, I am a postdoc in the DiverSE team in Rennes. I defended my PhD thesis entitled “Scaling Domain Agnostic Techniques for Program Synthesis” on 10/12/2024, I was supervised by Nathanaël Fijalkow in LaBRI. You can find my slides and my manuscript.

You can send me an email at theomatricon@gmail.com or theo.matricon@inria.fr.

Feel free to send me a mail just to chat about some interesting work, not necessarily yours!

Research Interests

My research so far lie in the domains of artificial intelligence, formal methods and software engineering. My profile lies at the intersection of these domains with the fascinating field of program synthesis. Here is a non exhaustive list of things I am excited about

• Program Synthesis, Code Generation
• Performance Variability
• Machine Learning tools that can help reduce environmental impact

Honestly everything’s interesting there’s just not enough time but I would like to emphasize that I am more interested in research that stands the test of time (for time = some years not a few months).

Quick info

Food intolerances: gluten, lactose, alcohol

Phone Number: my phone number

News

Current:

• Applying for “CR” positions
• June: Will be at the GDR GPL
• July: Will present our Enumeration algorithms at CNIA

2025:

• March 13: I will be presenting a talk at the ALMOST team seminar in Versailles
• March: I will be auditioned at CNRS in Jussieu for a research position
• March: Our paper on best first search for enumeration in constant delay was presented at AAAI 2025 by Guillaume Lagarde wit ha lof of interesting discussions.
• March: I visited Nathanaël Fijalkow with other members of the team such as Gabriel Bathie or Roman Kniazev.
• February: Our paper on replication of chess engine got accepted at IST joint work with Axel Martin, Mathieu Acher and Djamel Khelladi.
• February: Gave a talk at the GT VL & GLIA and attended VAMOS
• January: Here is a guide we wrote to CNRS applications with links to the other guides on the internet.
• January 24: I will give a talk at the LINKS seminar, at CRIStAL, in Lille on ranked enumeration of regular tree languages.
• January: I visited Simula in Oslo, Norway for a collaboration!

2024:

• December: Our paper on best first search for enumeration in constant delay got accepted at AAAI 2025 and was selected for an oral. Joint work with Nathanaël Fijalkow and Guillaume Lagarde
• December 10: I defended my PhD thesis! Yay~
• November: Gave a talk at the Synthesis Days in Bordeaux
• July 5: I was at the opening meeting of the LLM4Code project
• June 2-7: I was at the Dagsthul seminar on “Stochastic Games”
• June 17: I gave a talk on “Code Generation and Program Synthesis: where are we?” at a regional AI event
• June - July: Chaimaa Radiousse is working with me on estimating mixtures of stable distributions
• May - July: An intern Maxence is working on multi agent reinforcement learning in order to balance board games supervised by Guillaume Lagarde, Gabriel Bathie and I.
• March - July: Baptiste Mouillon is working on fast approaches for LTL learning with Nathanaël Fijalkow, Pierre Vandenhove, Gabriel Bathie and I.
• April 25-26: I was at GT DAAL in Rennes
• March-April: in Rennes from March 18 to April 26 to work with Mathieu Acher and the DiverSe team
• April 11: I gave a talk at 68nqrt seminar about “Scaling enumerative domain-agnostic program synthesis”
• March 28: I gave a talk at DiverSe Coffee
• February - Avril: Sylvain is working with Nathanaël Fijalkow, Pierre Vandenhove, Guillaume Lagarde and me on program synthesis for reinforcement learning
• Nov. 23 - March 24: Gianni Padula is working with Nathanaël Fijalkow, Pierre Vandenhove and me on a new reactive synthesis algorithm
• February: I gave a talk at Progress, a software engineering research team, on “Code Generation vs Program Synthesis”
• January: I took part in the organization and in the events of AI4Industry to talk about the PhD

2023:

• December: I went to the GDR GLIA in Paris to talk about WikiCoder
• November: With the other PhD Students in the psychosocial risks we’ve created a listening and speaking group for PhD students to be listened to in a safe space
• November: Gaëtan Margueritte is asking for more program synthesis and joined us as a research engineer until March.
• October: I invited Natacha Portier to talk about Impostor Syndrom to PhD Students
• September: I was at the Dagstuhl meeting on “The Futures of Reactive Synthesis” report.
• June: I am now mentoring Clara Marcille.
• June-July: I welcomed two interns:
  • 5th June to 29th July: Félix Yvonnet to work on restarts;
  • 19th June to 29th July: Priscilla Tissot to work on predictions.
• April 27: I presented WikiCoder: Learning to Write Knowledge-Powered Code at SPIN (ETAPS) in Paris.
• April 21: I presented our ongoing work on language sharpening at GT DAAL in Paris.
• March: I visited the Warsaw University (MIMUW) for two weeks.
• February: Our paper WikiCoder: Learning to Write Knowledge-Powered Code has been accepted at SPIN (ETAPS).
• January: Our paper Challenges of Acquiring Compositional Inductive Biases via Meta-Learning is finally published in the ECML/PKDD Workshop on Meta-Knowledge Transfer

2022:

• December: I am now finance officer of the local PhD association!
• October: I’m in Warsaw (MIMUW) for a visit!
• July: I gave a talk at GT Vérif in LaBRI on strategies as trees in MDPs.
• May - July: Utkarsh Rajan joined us for an internship on DeepSynth.
• March - June: Gaëtan Margueritte joined us for an internship on DeepSynth.
• February 25/26/27: I presented our DeepSynth paper at AAAI 2022 (Oral).
• February 11: Nathanaël Fijalkow has successfully defended his HDR.

2021:

• December 9: I participated in NAIA.R to mediate about AI.
• October 29, 30: Presentation at CP 2021 of our paper: Statistical Comparison of Algorithm Performance Through Instance Selection.
• October 2021: Now a PhD student at LaBRI.