Shivangi Tripathi

Shivangi Tripathi

Ph.D. Candidate in Computer Science, Texas State University

I am a Ph.D. candidate in Computer Science at Texas State University, where I work in the Human AI Synergy Lab with Dr. Heena Rathore.

My research is on hallucinations in generative AI: cases where a model produces output that sounds correct but is false or not supported by its input. I study this problem in language models that generate text, vision-language models that describe images, and audio-language models that process speech. I develop methods to detect and correct these errors, and I test how models fail under adversarial prompts, with applications in transportation safety and health care.

I will graduate in 2027 and am looking for full-time positions starting after graduation.

Outside of work, I go to the gym regularly and enjoy lifting weights. I also enjoy semi-classical Indian dance, especially Bharatanatyam, and I like hiking and playing badminton.

Shivangi Tripathi

Updates

Nov 2026Paper accepted to Findings of EACL 2027: “MS-CAPS: Minimal Sufficient Image Captions for Safety-Critical Traffic Scene Understanding.”
Oct 2026Taught two classes at Texas State University: Computer System Security (CS 4371), and Cryptography, where I covered security attacks on large language models.
Jun 2026Received a NAIRR Pilot compute award (12,400 GPU hours on NCSA Delta) for the project “Trustworthy Vision-Language Models via Minimal Sufficient Captions for Faithful Image Generation” (PI: Dr. Heena Rathore).
Apr 2026Defended my dissertation proposal and advanced to Ph.D. candidacy.
Mar 2026Taught a class in Computer System Security (CS 4371) on methods to detect and reduce hallucinations in large language models.
Sep 2025Taught a class in Computer System Security (CS 4371) at Texas State University.

All updates →

Selected Publications

  1. EACL 2027
    MS-CAPS: Minimal Sufficient Image Captions for Safety-Critical Traffic Scene Understanding
    Shivangi Tripathi, Heena Rathore
    Findings of the Association for Computational Linguistics (EACL), 2027
  2. CCNC 2026
    Paragraph-Level Hallucination Detection and Correction for Trustworthy Large Language Models in Networked Systems
    Shivangi Tripathi, Teancy Jennifer, Henry Griffith, Heena Rathore
    IEEE 23rd Consumer Communications & Networking Conference (CCNC), 2026
  3. AITest 2025
    REVISE: A Framework for Paragraph-Level Misinformation Correction in Large Language Models
    Shivangi Tripathi, Teancy Jennifer, Henry Griffith, Heena Rathore
    IEEE International Conference on Artificial Intelligence Testing (AITest), 2025

All publications →