About Me
Assistant Professor at West Virginia University, Lane Dept. of CS & EE
I study AI systems, especially conversational AI, like Large Language Models.
A recurring theme in my work is uncertainty, and how it can be leveraged for
- more accurate world models
- better human-AI collaboration
- understanding the limits of AI.
Previously, I competed my PhD at Northeastern University and my BS in Math at the University of Pittsburgh. I have also had the pleasure of collaborating with research teams at the Allen AI Institute, Amazon, and the Toronto Blue Jays baseball team.
Recent Good News
- Two outstanding paper awards [1,2] at EMNLP workshops
- Best Area Chair Award at EMNLP
- Top Reviewer Award at AISTATS
- Best Paper Award at UAI
Recent Research Directions
Language Models as Probabilistic World Models
Can language models provide calibrated uncertainty about their world knowledge?
[at ACL Findings 2024, ACL 2025]
Managing Uncertainty in Multi-Modal Interactions
Can language models effectively integrate perceptual information to communicate with humans?
[at EACL 2024, TACL 2024, EMNLP 2025, RANLP 2025]
Uncertainty for Bias Mitigation
Can language model uncertainty be used to mitigate biases, such as sycophancy?
Reasoning with Uncertainty
Can uncertainty be used to decompose complex tasks like coding, math, and embodied planning?
Anthony Sicilia 