About Me
Hi, I am Clément Romac. I am Research Scientist at Hugging Face and a last-year PhD student jointly supervised by Pierre-Yves Oudeyer (FLOWERS team, Inria Bordeaux) and Thomas Wolf (Hugging Face) studying how curiosity-driven RL can help grounding Large Language Models. My PhD is expected to finish around the end of 2025 and I am now looking for postdoctoral positions related to RL, curiosity-driven learning, or language grounding through embodied interactions. If you are interested, do not hesitate to reach out!
Previously, I worked as a Machine Learning Engineer at Weenove and as a Research Engineer in the FLOWERS team working on Automatic Curriculum Learning for Deep RL and Automated scientific discovery in complex systems (e.g. self-organizing). Apart from research, I have also been involved in various dissemination initiatives, in particular towards education (I am co-leading the development of the “ChatGPT explained in 5 minutes” series of videos that is used in French classrooms and now embedded in official trainings for workers from all French ministries). I am also the co-organizer of the Machine Learning Meetup group of Bordeaux.
Research Interests
My research focuses on Reinforcement Learning (RL) and how it can shape artificial agents’ knowledge. I notably study how Large Language Models (LLMs) can be grounded through online interactions with an environment using RL. I am interested in open-ended learning setups and my research often involves methods such as curiosity-driven RL or Automatic Curriculum Learning.
News
- [July 2025] We’re presenting MAGELLAN at ICML 2025 in Vancouver!
- [June 2025] We’re presenting our WorldLLM paper at RLDM 2025!
- [May 2025] I gave an invited talk at the ISIR lab from Sorbonne University on “Grounding LLMs through curiosity-driven online RL”.
- [May 2025] I gave a talk on generative AIs to La main à la pâte, a French association promoting science in classrooms.
- [March 2025] We just released our ICML 2025 MAGELLAN paper, a metacognitive framework for LLM agents in large goals spaces.
- [February 2025] Our new paper studying prompt overfitting in RL finetuning of LLMs got accepted at NAACL Findings 2025!
- [December 2024] Our SAC-GLAM paper got accepted at the IMOL workshop of NeurIPS 2024!
- [June 2024] I gave a talk at the French Ministry of Economics and Finance on “Explaining AI in 2024” for the LaborIA.
- [June 2024] Our JAT paper got accepted at the ARLET workshop of ICML 2024!
- [Mar. 2024] The english version of our “ChatGPT explained in 5 minutes” series is out!
- [Feb. 2024] Our large-audience article on grounding LLMs is on the front-page of Pour la Science February edition.
- [Feb. 2024] The sixth episode of our “ChatGPT expliqué en 5 minutes” series has been published.
Publications
-
RLDM
Guillaume Levy, Cedric Colas, Pierre-Yves Oudeyer, Thomas Carta, Clement Romac
Multi-disciplinary Conference on Reinforcement Learning and Decision Making (RLDM) 2025
-
ICML
Loris Gaven, Thomas Carta, Clément Romac, Sylvain Lamprier, Olivier Sigaud, Pierre-Yves Oudeyer
International Conference on Machine Learning (ICML), 2025.
-
NAACL Findings
Mohamed Salim Aissi, Clement Romac, Thomas Carta, Sylvain Lamprier, Pierre-Yves Oudeyer, Olivier Sigaud, Laure Soulier, Nicolas Thome
NAACL Findings 2025.
-
IMOL Workshop
Loris Gaven, Clément Romac, Thomas Carta, Sylvain Lamprier, Olivier Sigaud, Pierre-Yves Oudeyer
IMOL Workshop, NeurIPS 2024.
-
ARLET Workshop
Quentin Gallouédec, Edward Beeching, Clément Romac, Emmanuel Dellandrea
ARLET Workshop, ICML 2024.
-
ICML
Thomas Carta, Clément Romac, Thomas Wolf, Sylvain Lamprier, Olivier Sigaud, Pierre-Yves Oudeyer
International Conference on Machine Learning (ICML), 2023.
-
ICML
Clément Romac, Rémy Portelas, Katja Hofmann, Pierre-Yves Oudeyer
International Conference on Machine Learning (ICML), 2021.
-
Rémy Portelas, Clément Romac, Katja Hofmann, Pierre-Yves Oudeyer
Preprint, 2020
-
Clément Romac, Vincent Beraud
Preprint, 2019
Projects
-
ChatGPT explained in 5 minutes
A series of short videos explaining Large Language Models.
-
Lamorel
Lamorel is a Python library designed for RL practitioners eager to use Large Language Models (LLMs).
-
Automated Discovery Tool
A software for assisted and automated discovery of patterns in the exploration of complex systems.
-
Interactive Deep RL Demo
In-browser interactive demo of Deep Reinforcement Learning agents' adaptation to unknown 2D tasks.
Other stuff
- Teaching: You can find my teaching materials here.
- Music: I have been playing bass guitar for several years, including in a jazz fusion band
- Sport: In my spare time, I enjoy doing sport, from outdoor activities (hiking, mountaineering, running, trail, canoeing...) to football, tennis padel, badminton...
Powered by Jekyll and Minimal Light theme.