Scholar
Victor Wang
Google Scholar ID: 4izXmZYAAAAJ
University of Texas at Austin
Citations & Impact
Citations
37
H-index
4
i10-index
2
Publications
7
Co-authors
5
list available
Publications
Resume (English only)
Academic Achievements
- Calibrating Verbalized Confidence with Self-Generated Distractors, arXiv 2025 (first author)
- Improving LLM-as-a-Judge Inference with the Judgment Distribution, EMNLP 2025 Findings (first author)
- Designing LLM-Based Support for Homelessness Caseworkers, AAAI PubLLM 2024 (second author)
- Syllable-PBWT for space-efficient haplotype long-match query, Bioinformatics 2023 (first author)
- P-smoother: efficient PBWT smoothing of large haplotype panels, Bioinformatics Advances 2022 (third author)
- MorphSet: Improving Renal Histopathology Case Assessment Through Learned Prognostic Vectors, MICCAI 2021 (third author)
- Polaratio: A magnitude-contingent monotonic correlation metric and its improvements to scRNA-seq clustering, bioRxiv 2020 (first author)
Research Experience
- Summer 2022 Intern at Schlumberger (Menlo Park, CA): Developed an application enabling a robot to locate and approach sound sources using lidar, microphone arrays, and event-based cameras
- Course projects include:
- - CS 395T Topics in NLP (Fall 2023): Trained a Transformer LM to predict future token bags as a planning signal
- - CS 395T Grounded NLP (Spring 2024): Applied multimodal contrastive decoding to multilingual image captioning
- - CS 394R Reinforcement Learning (Spring 2024): Trained a Theory of Mind agent for the cooperative card game Hanabi
- - CS 378 NLP (Spring 2023): Enhanced paragraph segmentation with coreference-based attention
- - CS 378 Autonomous Driving (Fall 2023): Built a robot that interacts with humans to use elevators for multistory navigation
Background
- Final-year undergraduate student at UT Austin, majoring in Computer Science and Mathematics
- Member of the Turing Scholars honors program
- Research interests focus on natural language understanding and generation
- Interested in training objectives, model design, and inference procedures that enable controllable and versatile models
- Aims to build reliable, human-interfaceable models of people and the world
- Applying to PhD programs for Fall 2026
