Benjamin Guedj
Professor of Machine Learning and Foundational AI
Develops statistical learning theory, especially PAC-Bayesian generalisation bounds, for foundational AI.
Benjamin Guedj's lab at UCL works on learning theory & optimization and probabilistic & causal inference. 23 papers since 2023, mostly Neurocomputing. 3 PhD students. 28 alumni, most often to Senior Data Scientist at DiagRAMS Technologies.
At a glance
- papers, 2023–2026
- 6/ year
- at T1 venues
- 0%
- PhD students
- 3
- MS students
- 0
- alumni listed
- 28
Papers by yearfrom the lab homepage, 2023–2026
- 3 papers2026 · T1 0 · collecting
- 5 papers2025 · T1 0
- 5 papers2024 · T1 0
- 10 papers2023 · T1 0
T1 venuesOtherVenues: Neurocomputing
Research focuspaper titles that use each topic's words
- generalisation7 papers
- Learning theory & optimization6 papers
- Probabilistic & causal inference4 papers
- develops statistical0 papers
- especially0 papers
Paper list
From the lab homepage.
20263 papers
- Statistical Foundations and Applications of Artificial Intelligence, Machine Learning and Deep LearningChange of measure through the Legendre transform (opens in a new tab)
- Climate InformaticsCross-Domain Offshore Wind Power Forecasting: Transfer Learning Through Meteorological Clusters (opens in a new tab)
- Statistical Foundations and Applications of Artificial Intelligence, Machine Learning and Deep LearningFederated Learning with Nonvacuous Generalisation Bounds (opens in a new tab)
20255 papers
- International Conference on Algorithmic Learning Theory [ALT]A PAC-Bayesian Link Between Generalisation and Flat Minima (opens in a new tab)
- The Journal of Conventional Weapons DestructionAI for ERW Detection in Clearance Operations - The State of Research (opens in a new tab)
- International Conference on Algorithmic Learning Theory [ALT]Generalisation under gradient descent via deterministic PAC-Bayes (opens in a new tab)
- Foundations and Trends in Machine LearningGeneralization Bounds: Perspectives from Information Theory and PAC-Bayes (opens in a new tab)
- NeurIPS WorkshopInference and Generating Method for Extremely Sparse Networks
20245 papers
- Proceedings of The International Conference on Artificial Intelligence and Statistics [AISTATS]Comparing Comparators in Generalization Bounds (opens in a new tab)
- Advances in Neural Information Processing Systems [NeurIPS]Controlling Multiple Errors Simultaneously with a PAC-Bayes Bound (opens in a new tab)
- Physical Review EDiscord in the voter model for complex networks (opens in a new tab)
- StatGeneralisation Bounds for Kernel PCA through PAC-Bayes Learning (opens in a new tab)
- Advances in Neural Information Processing Systems [NeurIPS]Learning via Surrogate PAC-Bayes (opens in a new tab)
202310 papers
- Transactions on Machine Learning Research [TMLR]A note on regularised NTK dynamics with an application to PAC-Bayesian training (opens in a new tab)
- Bioresource TechnologyBayesian Uncertainty Quantification for Anaerobic Digestion models (opens in a new tab)
- Journal of Machine Learning Research [JMLR]Cluster-Specific Predictions with Multi-Task Gaussian Processes (opens in a new tab)
- Proceedings of the Paris Institute for Advanced Study, Intercontinental AcademiaCollective intelligence is needed to ensure beneficial Artificial Intelligence (opens in a new tab)
- Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems [NeurIPS]Learning via Wasserstein-Based High Probability Generalisation Bounds (opens in a new tab)
- Journal of Machine Learning Research [JMLR]MMD Aggregated Two-Sample Test (opens in a new tab)
- NeurocomputingModel Validation Using Mutated Training Labels: An Exploratory Study (opens in a new tab)
- Complex Networks and Their Applications XIOpening up echo chambers via optimal content recommendation (opens in a new tab)
- Transactions on Machine Learning Research [TMLR]PAC-Bayes Generalisation Bounds for Heavy-Tailed Losses through Supermartingales (opens in a new tab)
- Proceedings of The International Conference on Artificial Intelligence and Statistics [AISTATS]Tighter PAC-Bayes Generalisation Bounds by Leveraging Example Difficulty (opens in a new tab)
Where alumni went28 listed, 38% to industry
Industry 5Academia 8
Most often to
- Senior Data Scientist at DiagRAMS Technologies
Also to
- EPFL
- University of Oxford
- University of Tokyo
Apply to UCLFall 2027
Not announced yet
Computer Science MPhil/PhD
- Round
- Main
- Opens
- —
- GRE
- See program page
- English
- See program page
- Fee
- See program page
Not yet confirmed for this intake.
Similar labsnearest by research summary
- GBUCL
Tim Rocktäschel
- Reinforcement learning
- CNSJTU
Yue Ding
- Vision-language & multimodal
- CNPeking
Tiejun Huang
- Computational imaging & photography
- USGeorgia Tech
Bo Dai
- Reinforcement learning
- CAToronto
David Duvenaud
- Probabilistic & causal inference
- USMIT
Gabriele Farina
- Learning theory & optimization
Your notes
Save this lab to keep notes.
Data
Checked against the lab homepage.
- Collected
- Verified
- Papers
- The lab homepage
- Members
- Lab homepage
- Alumni
- Lab homepage
Tiers are venue classifications, not lab rankings.