Abstract: Transformer-based large language models (LLMs) have achieved remarkable success, yet many challenges remain. In this talk, I will address a fundamental question: Do all tokens require the same amount of computation within a Transformer? I will share insights into this question and introduce our dynamic layer-skipping and…

Abstract: Community detection is a long-standing problem with applications from social networks to biology. Given its popularity and that it is NP-complete, heuristics abound, though no gold standard exists. In this talk, I will introduce an entirely novel, graphlet-based algorithm for community detection: given an edge density…