Scholar
Siddartha Devic
Google Scholar ID: LVL-kmUAAAAJ
PhD Student, University of Southern California
Citations & Impact
Citations
133
H-index
6
i10-index
5
Publications
14
Co-authors
8
list available
Publications
Resume (English only)
Academic Achievements
- Publications: 'Trace Length is a Simple Uncertainty Signal in Reasoning Models', 'Auditability and the Landscape of Distance to Multicalibration', 'From Calibration to Collaboration: LLM Uncertainty Quantification Should Be More Human-Centered', 'When is Multicalibration Post-Processing Necessary?', 'An External Fairness Evaluation of LinkedIn Talent Search', 'Stability and Multigroup Fairness in Ranking with Uncertain Predictions', 'Fairness in Matching under Uncertainty', etc.
Research Experience
- Summer 2025: Research internship at Apple, Collaborators: Aravind Gollakota and Preetum Nakkiran; Fall 2024: Visitor at the Simons Institute, Participated in the Modern Paradigms in Generalization research program; Summer 2024: Research internship at Amazon, Collaborator: Gaurush Hiranandani.
Education
- PhD: University of Southern California (USC) Computer Science Department, Advisors: Vatsal Sharan and Aleksandra Korolova; Undergraduate: The University of Texas at Dallas (UTD), Majors: Math and Computer Science, Advisors: Nick Ruozzi, Jason Jue, Brendan Juba, Ben Raichel.
Background
- Research interests: trustworthy, fair, and robust machine learning and AI. Field: Computer Science.
Miscellany
- Received National Defense Science and Engineering Graduate (NDSEG, 2021-2024) Fellowship and USC Viterbi Graduate Fellowship (2024-2025).
