Dennis G. Wilson is a Professor at ISAE-Supaero of artificial intelligence and data science. His research is inspired by the many forms of biological intelligence, specifically through the study of evolutionary algorithms and neural networks. He earned his PhD at the Institut de Recherche en Informatique de Toulouse (IRIT), focused on the evolution of design principles for artificial neural networks. Prior to that, he worked on evolutionary developmental models in the Anyscale Learning For All group at CSAIL, MIT. His current work focuses on interpretable machine learning, exploration in search and learning, and environmental applications of AI, specifically how AI can help accelerate environmental science in the face of climate change.
A complete CV follows. It is also available as a PDF.
Academic Positions
- 2019-present: Professor at ISAE-Supaero
- Associate Professor from 2019 to 2025, HDR and tenure in 2025
- Professor of Artificial Intelligence and Data Science in the Learning, Decision, and Optimization group of the Complex Systems Engineering Department
- 2019: Postdoctoral researcher at University of Toulouse
- Research on the application of genetic programming to electrical grid crisis management
- 2016-2019: Lecturer at University of Toulouse
- Teaching Master’s classes in Machine Learning and Databases
- 2011-2014: Researcher at MIT Computer Science and Artificial Intelligence Lab, ALFA group
- Optimization using distributed evolutionary algorithms and genetic regulatory networks and data classification with genetic algorithms, advised by Kalyan Veeramachaneni and Una-May O’Reilly
Industry Experience
- 2019-2021: Co-Founder at Nautilia Computing, Toulouse, France
- Artificial Intelligence consulting and services for environment observation and simulation
- 2014-2016: Software engineer at Infinidat LTD, Israel
- Full-time position in infrastructure development for data storage systems
Education
- 2025: HDR in Computer Science (Habilitation à diriger des recherches)
- “Methods in Exploration and Interpretability for Automating Discovery”
- The highest academic degree in France, allowing for research direction
- Université Toulouse Capitole, France
- 2016-2019: PhD in Computer Science, IRIT, University of Toulouse, France
- “Evolving Principles of Artificial Neural Design”
- Director: Prof. Hervé Luga, Université Toulouse - Jean Jaurès, IRIT
- Co-supervisor: Prof. Sylvain Cussat-Blanc, Université Toulouse Capitole, IRIT
- 2010-2014: BSci in Electrical Engineering and Computer Science, MIT, Cambridge, USA
Teaching
- 2020-present: Decision and Data Science, ISAE-Supaero
- Leading this Master’s level program covering machine learning, data engineering, and applications of AI over 240 hours
- Created new courses on deep learning, data storage, continuous and combinatorial optimization, cloud orchestration, and legal frameworks around AI
- Program has roughly 60 students per year, with alumni now in positions at top organizations like Amazon, HuggingFace, Mistral, and the European Commission on AI
- 2020-present: Evolutionary Algorithms, ISAE-Supaero
- Teaching this class on evolutionary algorithms, focusing on policy search in a project on evolution of soft robots
- Created the material for this class covering evolutionary algorithms, genetic algorithms, evolutionary strategies, neuroevolution, genetic programming, and quality diversity
- 2018-2019: Computational Intelligence, Lecturer, University of Toulouse
- 2016-2018: Databases, Lecturer, University of Toulouse
- 2014: Introduction to Python, Teaching Assistant, MIT
Advising
I have co-advised 6 PhD students, 4 of which have now successfully defended. I have fully supervised one postdoctoral researcher, Erwan Lecarpentier, and am currently co-supervising the postdoctoral research of Giorgia Nadizar. I have supervised 11 Master’s students for their thesis projects and have served as academic advisor for approximately 10 Master’s level final internships per year since 2020.
Current PhD Students
- 2023-present: Camilo de la Torre
- PhD in hybridization between Cartesian Genetic Programming and specialized machine learning
- Advisors: Sylvain Cussat-Blanc, Dennis G. Wilson, Hervé Luga
- Financing: EDMITT scholarship
- 2023-present: Estelle Chigot
- PhD in synthetic-to-real domain adaptation for object recognition
- Advisors: Thomas Oberlin, Dennis G. Wilson, Meriem Ghrib
- Financing: CIFRE with Airbus
Completed PhD Students
- 2022-2025: Paul Antoine le Tolguenec
- Exploration Methods for Reinforcement Learning Applied to Critical System Testing
- Advisors: Emmanuel Rachelson, Dennis G. Wilson, Yann Besse
- Financing: CIFRE ANITI with Airbus
- Defended May 7, 2025
- 2021-2024: Paul Templier
- Leveraging Structure in Evolutionary Neural Policy Search
- Advisors: Emmanuel Rachelson, Dennis G. Wilson
- Financing: Half-scholarships from ISAE-Supaero and Région Midi-Pyrénées
- Defended April 22, 2024
- 2020-2023: Mahmoud Al-Najar
- Modelling coastal evolution with machine learning
- Advisors: Rafael Almar, Jean-Marc Delvit, Dennis G. Wilson
- Financing: Half-scholarships from CNES and Région Midi-Pyrénées
- Defended November 30, 2023
- 2020-2023: Kaitlin Maile
- Dynamic Architectural Optimization of Artificial Neural Networks
- Advisors: Hervé Luga, Sylvain Cussat-Blanc, Dennis G. Wilson
- Financing: EDMITT Scholarship
- Defended October 4, 2023
Grants, Honors & Awards
- 2025: Best Paper Award - GECCO Evolutionary Machine Learning Track
- For “Evolution of Inherently Interpretable Visual Control Policies”
- 2025: Best Student Paper Award - IJCCI
- For “Extending Cartesian Genetic Programming via Iterative Subgraph Assessment”
- 2025: Habilitation à diriger des recherches
- Approved by an international jury of experts in machine learning, evolutionary computation, and environmental data science
- Validates research direction capabilities and allows advising PhD students independently
- 2024: ACM SIGEVO Human Competitive Competition Gold Award
- First place in the “Humies” Competition at GECCO 2024 for work on interpretable image analysis
- Recognizes results that are competitive with human performance
- 2024: Best Paper Award - GECCO Complex Systems Track
- For “Quality with Just Enough Diversity in Evolutionary Policy Search”
- 2024: ACM SIGSOFT Distinguished Paper Award
- At the International Symposium on Software Testing and Analysis for “Exploration-Driven Reinforcement Learning for Avionic System Fault Detection”
- 2024: Best Paper Award - EuroGP
- For “Naturally Interpretable Control Policies via Graph-Based Genetic Programming”
- 2024-present: ANITI Affiliate Member
- Associated with the Artificial and Natural Intelligence Toulouse Institute (ANITI)
- Awarded status of Affiliate Member for contributions on interpretable machine learning
- 2022: CIFRE thesis grant with Airbus
- For Paul-Antoine le Tolguenec’s PhD
- 2021-2025: Two CIFRE thesis grants
- Industrial grants for collaborations with Airbus on AI research applied to aerospace engineering
- 2021-2023: Invited Speaker on Evolutionary Reinforcement Learning
- 2020: Winner, GECCO competition on agent control in DOTA
- 2020: ACM SIGEVO Best Dissertation Award
- Yearly award recognizing the best doctoral dissertation in evolutionary computation
- 2019-2023: Région Occitanie grants
- Two thesis grants for research on combining evolutionary algorithms and machine learning, and on applying machine learning to coastal science
- 2019: Thesis grant from the CNES on shoreline forecasting
- 2019: Thesis grant from La Région Occitanie and ISAE-Supaero on neuroevolution
- 2017: SIGEVO student representative, ACM Turing Award Celebration
- 2017: Winner, ACM SIGAI Essay Contest on Ethics and AI
- For essay on “The ethics of automated behavioral microtargeting”
- 2015: CIMI Doctoral Fellowship recipient, France
- 2013: Best paper nomination, GECCO 2013 GDS track
Research Focus
My work spans evolutionary algorithms, machine learning, and the application of AI to climate science. I pursue three main research theses: genetic programming offers an interpretable and competitive alternative to deep learning; exploration in search and learning can improve AI robustness; and machine learning can help us understand climate change.
Graph-based Genetic Programming
My research demonstrates that Graph-based Genetic Programming (GraphGP) can generate interpretable models competitive with state-of-the-art deep learning. Key publications include:
- Evolving Simple Programs for Playing Atari Games (GECCO 2018): Demonstrated GraphGP’s competitiveness with deep reinforcement learning on the Atari benchmark, laying the foundation for interpretable genetic programming
- Naturally Interpretable Control Policies (EuroGP 2024, Best Paper Award): Showed GraphGP creates interpretable control policies rivaling deep reinforcement learning for standard robotic tasks
- Evolutionary Design of Explainable Algorithms for Biomedical Image Segmentation (Nature Communications 2023, Humies Gold Award): Presented a GraphGP method for optimizing interpretable analysis pipelines, achieving performance competitive with state-of-the-art neural networks
- Evolution of Inherently Interpretable Visual Control Policies (GECCO 2025, Best Paper Award): Demonstrated that interpretable policies can solve visual control tasks at performance similar to deep reinforcement learning while remaining fully transparent
Exploration in Search and Learning
My work advances exploration techniques in both evolutionary algorithms and reinforcement learning:
- Quality with Just Enough Diversity in Evolutionary Policy Search (GECCO 2024, Best Paper Award): Introduced a novel evolutionary policy search algorithm balancing exploration and exploitation for more robust and diverse policy generation
- Exploration-Driven Reinforcement Learning for Avionic System Fault Detection (ISSTA 2024, SIGSOFT Distinguished Paper Award): Demonstrated that evolutionary exploration algorithms can detect faults in critical avionic systems not found by standard verification methods
- Exploration by Learning Diverse Skills through Successor State Measures (NeurIPS 2024): Presented LEADS, an exploration algorithm leveraging successor state measures to learn diverse skills and enhance exploration in multi-task environments
Machine Learning for Climate Science
I apply machine learning to accelerate environmental science in the face of climate change:
- Satellite Derived Bathymetry Using Deep Learning (Machine Learning 2021): Introduced a deep learning model for estimating bathymetry from satellite data, improving coastal geography analysis
- Improving a Shoreline Forecasting Model with Symbolic Regression (ICLR CCAI Workshop 2023): Used genetic programming to improve existing shoreline models, demonstrating significant improvements in forecasting accuracy
- Predicting beach profiles with machine learning from offshore wave reflection spectra (Environmental Modelling & Software 2025): Detailed a machine learning approach to forecasting beach profiles using wave reflection data
Invited Talks
- 2025: Series of 5 lectures on AI in society, Quint Fonsegrives
- 2023: Impact of AI on Education, ANITI 2023
- 2023: Keynote, Evolutionary Reinforcement Learning Workshop, GECCO 2023
- 2021: Evolutionary Reinforcement Learning, Reinforcement Learning Virtual School
- 2018: Julia for Machine Learning, ANF APSEM 2018, CNRS, Toulouse
- 2018: Evolving simple programs for playing Atari games, IT University of Copenhagen
- 2017: Evolving neural programs for continuous learning, CSAIL, MIT
Contributions to the Research Community
Editorial Activities
- present: Editorial board member, ACM Transactions on Evolutionary Learning and Optimization
- Led the journal’s policy on the use of large language models
- Organized a special issue on the intersection of evolutionary computation and Large Language Models
- 2025-2027: Track Chair, Neuroevolution Track, GECCO
- Two-year mandate similar to Area Chair
- 2020-2022: Track Chair, Complex Systems Track, GECCO
- present: Regular reviewer for top AI conferences and journals
- NeurIPS, ICLR, ICML, AAAI, IJCAI, GECCO, IEEE Transactions on Evolutionary Computation
Workshop and Conference Organization
- 2026: Co-Local Chair, Evostar 2026
- Leading European event on bio-inspired AI, to be held in Toulouse in April 2026
- 2023-present: Founder and organizer, Graph-based Genetic Programming workshop
- Hosted at international conferences (GECCO, ALIFE, PPSN)
- 2024: Organizer, local workshop on Evolutionary Machine Learning, Toulouse
- 2018-2020: Organizer, Developmental Neural Network workshop
Competition Organization
- 2024-present: Co-organizer, Interpretable Control Policies competition at GECCO
- 2014-2016: Organizer, Wind Farm Layout Optimization competition at GECCO
- Culminated in an article in Renewable Energy
Diversity and Inclusion
- 2022-2024: Co-organizer, ANITI Diversity Commission
- 2021-2022: Secretary, Diversity, Equity, and Inclusion Committee, International Society of Artificial Life
AI for Climate
- 2021-present: Mentor, Climate Change AI Mentorship program
- Supporting early-stage researchers in AI and climate science
- 2021-2023: Member, AI for the Environment Committee (ENVIA)
- Promoting collaboration between AI researchers and environmental scientists
- 2022-2024: Faculty representative, Horizons committee at ISAE-Supaero
- Promoting sustainable development
- Organized special issue in Remote Sensing Journal on integrating satellite remote sensing with AI for coastal issues
- Organized special session at IEEE World Congress on Computational Intelligence on “AI for Climate Science”
Public Outreach
- Invited speaker on the impact of generative AI on education
- Invited speaker on the impact of AI on society
- Invited speaker on Large Language Models
- Maintained a newsletter with over 130 subscribers on recent trends in and societal impacts of AI
Personal
My main hobbies are reading, hiking, cooking, and games of all sorts. I am currently developing a tabletop role-playing game and enjoy challenging video games. I am a level 2 scuba diver in the French system and would love to dive with whales someday. I speak English natively, French fluently (C1), Hebrew conversationally, a fair bit of Spanish, and am starting to learn Japanese.