Skip Navigation Logo for: The Connecticut Conference of Independent Colleges
2nd Annual AI Research Symposium at Hartford AI Day

Showcasing innovative AI research across Connecticut.
Co-organized by the Connecticut
AI Alliance (CAIA) and the University of Connecticut (UConn).

 Date: Wednesday, April 22, 2026

 Venue: 100 Constitution Plaza, Hartford, CT

 Rooms: GBLC Room 501 (Track 1) and GBLC Room 502 (Track 2)

 Format: 15-minute presentations + 5-minute Q&A

 Full Hartford AI Day schedule: Track 2

CAIA invited undergraduate students, graduate students, and faculty members from all Connecticut colleges and universities to submit one-page abstracts of their AI-related research for presentation at the 2nd Annual AI Research Symposium, held as part of Hartford AI Day 2026

The symposium welcomed submissions based on recently published work, early results from ongoing research, or work currently under review or accepted at other venues. This non-archival format allowed researchers to share their latest findings while maintaining the ability to publish elsewhere.

Building on the success of our inaugural 2025 symposium, this year's event continues to bridge academic research with industry applications and foster collaboration across institutions statewide.

 

Accepted Papers & Presentation Schedule

Session 1: 9:45 - 10:30 AM

Track 1 (Room 501)

1) A Foundation Model for Microbiome Data Using Transformer Architectures
   Authors: Khaled Sayed, Mohit Raiyani
   Affiliation: University of New Haven

Track 2 (Room 502)

2) FedAVOT: Exact Distribution Alignment in Federated Learning via Masked Optimal Transport
   Authors: Herlock Rahimi, Dionysis Kalogerias
   Affiliation: Yale University
   Paper: arXiv:2509.14444

3) Toxic by Design: Psychophysiological Evidence That AI Personality Shapes Cognition,
     Emotion, and Work Quality
    Authors: Tamilla Triantoro, Leon Ciechanowski, Anna Kovbasiuk, Konrad Sowa, Aleksandra
    Przegalinska, Richard Freeman
    Affiliations: Quinnipiac University; Kozminski University; MIT Center for Collective Intelligence;
    Harvard University; NBER

 


Session 2: 11:15 AM - 12:15 PM

Track 2 (Room 502)

4) Consistency Is Not Safety: Paraphrase Sensitivity Reveals Text-Driven Selection in Medical
     Vision-Language Models

    Authors: Binesh Sadanandan, Vahid Behzadan
    Affiliation: University of New Haven
    Paper: arXiv:2603.20985

5) From Convolutions to Attention: Comparing CNN and Transformer Models for Medical Imaging
    Authors: Sarah Tasneem, Kehan Gao
    Affiliation: Eastern Connecticut State University

 


Session 3: 1:45 - 2:30 PM

Track 1 (Room 501)

6) Dynamic Automated Risk Mapping for Archaeological Looting
    Authors: Michelle D. Fabiani
    Affiliation: University of New Haven

Track 2 (Room 502)

7) AI Tutoring in Computer Science Education  
    Authors: Joel Rosiene, Carolyn Pe Rosiene
    Affiliations: Eastern Connecticut State University; University of Hartford
    Paper: IEEE Xplore

8) Teaching Symbolic AI through Ontologies, Linked Data and Knowledge Graphs
    Authors: Neli Zlatareva
    Affiliation: Central Connecticut State University

9) Working With AI: The Trust Verification Paradox in Human-AI Collaboration
    Authors: Beth Kanyiri
    Affiliation: Quinnipiac University

 


Session 4: 3:30 - 4:30 PM

Track 1 (Room 501)

10) Robust Coordination under Misaligned Communication via Power Regularization
     Authors: Nancirose Piazza, Amirhossein Karimi, Behnia Soleymani, Vahid Behzadan, Stefan Sarkadi
     Affiliations: University of New Haven; University of Tehran; King's College London
     Paper: IOS Press

12) Model Forensics: Characterizing Nondeterminism in Locally Hosted Quantized Language
       Models across Heterogeneous Inference Environments

      Authors: Sadie Simek, Robert Ruiz, Jacob Mancini, Ramyapandian Vijayakanthan
      Affiliation: Central Connecticut State University

Track 2 (Room 502)

13) HazDepth: A Fused Depth-Aware Detection Pipeline for Real-Time Hazard Recognition
     Authors: Shruti Brahma, Maral Talebbeidokhti, Khaled Sayed
     Affiliation: University of New Haven

14) Generating Quality Word-Association Puzzles
     Authors: Ashish Khadka, Mohamad Nassar, Muhammad Aminul Islam
     Affiliation: University of New Haven

15) Explainable RAG-based Agentic System for Conversational Scams Detection
     Authors: Ahmed Omar Salim Adnan, Yogananda Manjunath, Shivanjali Khare
     Affiliation: University of New Haven

16) Financial Literacy as a Driver of Financial Innovation Adoption: Evidence from
       AI-Based Advice and Cryptocurrency Holdings
      Author: Li Liang
      Affiliation: Eastern Connecticut State University