News
Recent milestones. The full archive is a click away.
"Muting tokens: exploring token sparsification for efficient audio transformers" is now on arXiv.
"Coresets are more than replay: a data-centric view of continual learning" published in Neural Computing and Applications.
"Unlocking [CLS] features for continual post-training" accepted to TMLR.
"Automated machine learning for unsupervised tabular tasks" accepted to MLJ.
"Pruned adaptation modules" accepted to CPAL.
"Self-regulated neurogenesis for online data-incremental learning" accepted for an oral presentation at CoLLAs.
Joined the General Assembly of the EU project SYNERGIES in Paris, France.
Participated in the "Learning Over Time" spring school at Certosa di Pontignano, Italy.
Attended the GeMMS summer school on generative modelling at TU Eindhoven.
"Continual learning with dynamic sparse training" accepted for an oral presentation at CPAL.
"AdaCL: adaptive continual learning" accepted for an oral presentation at the ContinualAI Unconference.
"Condition monitoring and predictive maintenance in railways" granted an international patent.
AutoML for hole-selective layers in perovskite solar cells published in Energy Technology.
"Predicting perovskite bandgap and solar cell performance with machine learning" published in Solar RRL.
ML for metal-oxide polymer composites in perovskite solar cells published in ChemPlusChem.
Comparison of ML models on electrochromic devices published in ACS Omega.
Publications
Recent work on continual learning and efficient model adaptation, along with earlier work on ML for materials science.
Energy Technology 2023
Automated machine learning in material discovery of hole selective layers for perovskite solar cells
ChemPlusChem 2021
A machine learning approach for metal oxide based polymer composites as charge selective layers in perovskite solar cells
ACS Omega 2020
Comparison of machine learning models on performance of single- and dual-type electrochromic devices
Supervision
If you are excited about models that keep learning, from continual learning and efficient model adaptation to systems that operate in changing real-world environments, I would love to brainstorm ideas and support your journey in a collaborative, low-pressure environment. I have been lucky to work with these talented and curious people:
Casper Adolfse
Memory for continual social-physical world models
Michael Rayan
Marker-free vision-based autonomous UAV precision landing
Danushkumar Venkadesh
Progressive orthogonal subspace consolidation for continual foundation model adaptation
Shreya Sajid
Muting tokens: exploring token sparsification for efficient audio transformers
Athira Kulampurath
Reliability and uncertainty in foundation model-based class-incremental learning
Coen Schouten
On-the-fly curriculum design for online class-incremental learning