Ransalu Senanayake
Assistant Professor
Computer Science

Welcome to the Laboratory for Learning Evaluation of autoNomous Systems (LENS Lab). We are interested in operationalizing machine learning (ML) models in autonomous systems and robots. Our objective is to develop a range of algorithmic tools for evaluating ML models utilized in such systems. We firmly believe that assessing models solely based on accuracy on a held-out dataset is insufficient for successful operationalization. Therefore, we delve deeper into critical questions: How can we explain black-box neural networks? How can we evaluate uncertainty of decisions? When and how do models fail? Are models fair and unbiased? Will a model perform effectively in new or changing environments? We examine these questions before, during, and after deployment. These efforts not only aid engineers in debugging models but also assist legislative bodies in establishing legal and ethical guidelines.



News

Dec '24
Som and Aditya will be presenting four workshop papers at NeurIPS'24.
Nov '24
We are currently accepting PhD students with strong backgrounds in ML, robotics, CV, or NLP. Master's and undergrad students enrolled at ASU can also apply to the LENS Lab. See Join Us!
Nov '24
Kelly Raines defended her Barrett Honors thesis titled "Generative AI-Aided Navigation for the Visually Impaired and Blind". Congrats!
May '24
Som's paper "Failures Are Fated, But Can Be Faded: Characterizing and Mitigating Unwanted Behaviors in Large-Scale Vision and Language Models" was accepted to ICML'24 as a spotlight-designated paper 🎗 (top 3.5%).