← Search

Hector Ignacio Jimenez Orellana

2 accepted papers

2025

Continuity and Isolation Lead to Doubts or Dilemmas in Large Language Models

NeurIPS 2025poster

Understanding how Transformers work and how they process information is key to the theoretical and empirical advancement of these machines. In this work, we demonstrate the existence of two phenomena in Transformers, namely _isolation_ and _continuity_. Both of these phenomena hinder Transformers t…

Cited by 2SourceScholar
2025

Strassen Attention, Split VC Dimension and Compositionality in Transformers

NeurIPS 2025poster

We propose the first method to show theoretical limitations for one-layer softmax transformers with arbitrarily many precision bits (even infinite). We establish those limitations for three tasks that require advanced reasoning. The first task, Match 3 (Sanford et al., 2023), requires looking at al…

Cited by 0SourceScholar