Alert button
Picture for Sotiris Anagnostidis

Sotiris Anagnostidis

Alert button

ETH Zurich

Towards Meta-Pruning via Optimal Transport

Feb 13, 2024
Alexander Theus, Olin Geimer, Friedrich Wicke, Thomas Hofmann, Sotiris Anagnostidis, Sidak Pal Singh

Viaarxiv icon

Harnessing Synthetic Datasets: The Role of Shape Bias in Deep Neural Network Generalization

Nov 10, 2023
Elior Benarous, Sotiris Anagnostidis, Luca Biggio, Thomas Hofmann

Figure 1 for Harnessing Synthetic Datasets: The Role of Shape Bias in Deep Neural Network Generalization
Figure 2 for Harnessing Synthetic Datasets: The Role of Shape Bias in Deep Neural Network Generalization
Figure 3 for Harnessing Synthetic Datasets: The Role of Shape Bias in Deep Neural Network Generalization
Figure 4 for Harnessing Synthetic Datasets: The Role of Shape Bias in Deep Neural Network Generalization
Viaarxiv icon

Navigating Scaling Laws: Accelerating Vision Transformer's Training via Adaptive Strategies

Nov 06, 2023
Sotiris Anagnostidis, Gregor Bachmann, Thomas Hofmann

Figure 1 for Navigating Scaling Laws: Accelerating Vision Transformer's Training via Adaptive Strategies
Figure 2 for Navigating Scaling Laws: Accelerating Vision Transformer's Training via Adaptive Strategies
Figure 3 for Navigating Scaling Laws: Accelerating Vision Transformer's Training via Adaptive Strategies
Figure 4 for Navigating Scaling Laws: Accelerating Vision Transformer's Training via Adaptive Strategies
Viaarxiv icon

Transformer Fusion with Optimal Transport

Oct 15, 2023
Moritz Imfeld, Jacopo Graldi, Marco Giordano, Thomas Hofmann, Sotiris Anagnostidis, Sidak Pal Singh

Figure 1 for Transformer Fusion with Optimal Transport
Figure 2 for Transformer Fusion with Optimal Transport
Figure 3 for Transformer Fusion with Optimal Transport
Figure 4 for Transformer Fusion with Optimal Transport
Viaarxiv icon

Scaling MLPs: A Tale of Inductive Bias

Jun 23, 2023
Gregor Bachmann, Sotiris Anagnostidis, Thomas Hofmann

Figure 1 for Scaling MLPs: A Tale of Inductive Bias
Figure 2 for Scaling MLPs: A Tale of Inductive Bias
Figure 3 for Scaling MLPs: A Tale of Inductive Bias
Figure 4 for Scaling MLPs: A Tale of Inductive Bias
Viaarxiv icon

Dynamic Context Pruning for Efficient and Interpretable Autoregressive Transformers

May 25, 2023
Sotiris Anagnostidis, Dario Pavllo, Luca Biggio, Lorenzo Noci, Aurelien Lucchi, Thomas Hoffmann

Figure 1 for Dynamic Context Pruning for Efficient and Interpretable Autoregressive Transformers
Figure 2 for Dynamic Context Pruning for Efficient and Interpretable Autoregressive Transformers
Figure 3 for Dynamic Context Pruning for Efficient and Interpretable Autoregressive Transformers
Figure 4 for Dynamic Context Pruning for Efficient and Interpretable Autoregressive Transformers
Viaarxiv icon

OpenAssistant Conversations -- Democratizing Large Language Model Alignment

Apr 14, 2023
Andreas Köpf, Yannic Kilcher, Dimitri von Rütte, Sotiris Anagnostidis, Zhi-Rui Tam, Keith Stevens, Abdullah Barhoum, Nguyen Minh Duc, Oliver Stanley, Richárd Nagyfi, Shahul ES, Sameer Suri, David Glushkov, Arnav Dantuluri, Andrew Maguire, Christoph Schuhmann, Huu Nguyen, Alexander Mattick

Figure 1 for OpenAssistant Conversations -- Democratizing Large Language Model Alignment
Figure 2 for OpenAssistant Conversations -- Democratizing Large Language Model Alignment
Figure 3 for OpenAssistant Conversations -- Democratizing Large Language Model Alignment
Figure 4 for OpenAssistant Conversations -- Democratizing Large Language Model Alignment
Viaarxiv icon

Random Teachers are Good Teachers

Feb 23, 2023
Felix Sarnthein, Gregor Bachmann, Sotiris Anagnostidis, Thomas Hofmann

Figure 1 for Random Teachers are Good Teachers
Figure 2 for Random Teachers are Good Teachers
Figure 3 for Random Teachers are Good Teachers
Figure 4 for Random Teachers are Good Teachers
Viaarxiv icon

Cosmology from Galaxy Redshift Surveys with PointNet

Nov 22, 2022
Sotiris Anagnostidis, Arne Thomsen, Tomasz Kacprzak, Tilman Tröster, Luca Biggio, Alexandre Refregier, Thomas Hofmann

Figure 1 for Cosmology from Galaxy Redshift Surveys with PointNet
Figure 2 for Cosmology from Galaxy Redshift Surveys with PointNet
Figure 3 for Cosmology from Galaxy Redshift Surveys with PointNet
Figure 4 for Cosmology from Galaxy Redshift Surveys with PointNet
Viaarxiv icon