Publications | Jannik Kossen

In-Context Learning in Large Language Models Learns Label Relationships but Is Not Conventional Learning.

International Conference on Learning Representations (ICML). 2024.

PDF arXiv

Three Towers: Flexible Contrastive Learning with Pretrained Image Models.

Jannik Kossen, Mark Collier, Basil Mustafa, Wang Xia, Xiaohua Zhai, Lucas Beyer, Andreas Steiner, Jesse Berent, Rodolphe Jenatton, Efi Kokiopoulou

Advances in Neural Information Processing Systems (NeurIPS) (Previously at Workshop on Efficient Systems for Foundation Models at ICML). 2023.

PDF arXiv Workshop

Active Acquisition for Multimodal Temporal Data: A Challenging Decision-Making Task.

Jannik Kossen, Cătălina Cangea, Eszter Vértes, Andrew Jaegle, Viorica Patraucean, Ira Ktena, Nenad Tomasev, Danielle Belgrave

Transactions on Machine Learning Research (TMLR) (Previously at Foundation Models for Decision Making Workshop at NeurIPS 2022). 2023.

PDF Poster Slides Video arXiv OpenReview Workshop

Active Surrogate Estimators: An Active Learning Approach to Label-Efficient Model Evaluation.

Jannik Kossen, Sebastian Farquhar, Yarin Gal, Tom Rainforth

Advances in Neural Information Processing Systems (NeurIPS Oral). 2022.

PDF Code Poster Slides Video arXiv

Self-Attention Between Datapoints: Going Beyond Individual Input-Output Pairs in Deep Learning.

Jannik Kossen, Neil Band, Clare Lyle, Aidan N Gomez, Tom Rainforth, Yarin Gal

Advances in Neural Information Processing Systems (NeurIPS). 2021.

PDF Code Slides Video arXiv

Active Testing: Sample-Efficient Model Evaluation.

Jannik Kossen, Sebastian Farquhar, Yarin Gal, Tom Rainforth

International Conference on Machine Learning (ICML). 2021.

PDF Code Poster Slides Video arXiv

Structured object-aware physics prediction for video modeling and planning.

Jannik Kossen, Karl Stelzner, Marcel Hussing, Claas Voelcker, Kristian Kersting

International Conference on Learning Representations (ICLR). 2020.

PDF Code arXiv

Modelling Videos of Physically Interacting Objects.

Jannik Kossen

Master Thesis (Heidelberg University). 2020.

PDF

Wie Maschinen lernen.

Kristian Kersting, Christoph Lampert, Constantin Rothkopf (Editors)

Springer. (Edited Volume. Jannik is author of 11 chapters.). 2019.

Springer SpringerLink Amazon Thalia