ScaDS.AI announces and welcomes you to join our public colloquium session on Monday, May 15, 2023 at 15:15-16:45 CEST. The colloquium takes place at ScaDS.AI Leipzig, Humboldtstr. 25, 04105 Leipzig, in the large seminar room (A03.07 Zwenkauer See) and in parallel online (link to Zoom session).
Associate Professor, Scientific Computing and Imaging (SCI) Institute, University of Utah, USA
Topology of Artificial Neuron Activations in Deep Learning: From Images to Word Embeddings
Deep convolutional neural networks such as GoogLeNet and ResNet have become ubiquitous in image classification tasks, whereas transformer-based language models such as BERT and its variants have found widespread use in natural language processing. In this talk, I will discuss recent efforts in exploring the topology of artificial neuron activations in deep learning, from images to word embeddings.
First, I will discuss the topology of convolutional neural network activations, which provides semantic insight into how these models organize hierarchical class knowledge at each layer. Second, I will discuss the topology of word embeddings from transformer-based models. I will explore the topological changes of word embeddings during the fine-tuning process of various models and discover model confusions in the embedding spaces. If time permits, I will discuss on-going work in studying the topology of neural activations under adversarial attacks.
Prof. Dr. Bei Wang Phillips
- Works on combining topological, geometric, statistical, data mining, and machine learning techniques with visualization for information exploration and scientific discovery
- PhD in Computer Science at Duke University, USA, in 2010
- Recent awards: DOE Early Career Research Program in 2020, NSF CAREER in 2022
Location
ScaDS.AI Leipzig
Löhrs Carré
Humboldtstraße 25
3rd floor
04105 Leipzig
Check out our event calendar for more information about our upcoming events!