5 papers
Topological Signatures of Grokking
Yifan Tang, Qiquan Wang, Inés GarcÃa-Redondo +1
We study the grokking phenomenon through the lens of topology. Using persistent homology on point clouds derived from the embedding matrices of a range of models trained on modular…
The Shape of Adversarial Influence: Characterizing LLM Latent Spaces with Persistent Homology
Aideen Fay, Inés GarcÃa-Redondo, Qiquan Wang +2
Existing interpretability methods for Large Language Models (LLMs) predominantly capture linear directions or isolated features. This overlooks the high-dimensional, relational, an…
A Topological Gaussian Mixture Model for Bone Marrow Morphology in Leukaemia
Qiquan Wang, Anna Song, Antoniana Batsivari +2
Acute myeloid leukaemia (AML) is a type of blood and bone marrow cancer characterized by the proliferation of abnormal clonal haematopoietic cells in the bone marrow leading to bon…
Confidence Bands for Multiparameter Persistence Landscapes
Inés GarcÃa-Redondo, Anthea Monod, Qiquan Wang
Multiparameter persistent homology is a generalization of classical persistent homology, a central and widely-used methodology from topological data analysis, which takes into acco…
On the Limitations of Fractal Dimension as a Measure of Generalization
Charlie B. Tan, Inés GarcÃa-Redondo, Qiquan Wang +2
Bounding and predicting the generalization gap of overparameterized neural networks remains a central open problem in theoretical machine learning. There is a recent and growing bo…