Yuhui Zhang
PhD Candidate
Palo Alto, California, United States
Summary
🤩
Rockstar🎓
Top SchoolYuhui Zhang is a Stanford CS PhD candidate and software engineer with a decade of experience building and evaluating NLP and ML systems from Palo Alto. She has contributed core back-end and ML engineering work to widely used open-source projects like Stanford's Stanza, refactoring data structures and implementing POS tagging, lemmatization, and dependency parsing components. Her research and engineering blend is evident in extensions to the HELM evaluation framework, where she integrated multiple commonsense QA benchmarks and improved calibrated probability computations for robust model assessment. Comfortable across production code and evaluation pipelines, she brings rigorous testing and careful interface design to complex language-modeling stacks. An uncommon strength is her ability to translate research-grade models into maintainable, well-tested engineering artifacts used by both researchers and practitioners.
10 years of coding experience
4.0, 4.0 at Stanford University
Github Skills (16)
neural-network10
pytorch10
dependency-parsing10
machine-learning10
lemmatization10
nlp10
deep-learning10
python10
natural-language-processing10
ml10
data-structure9
algorithm9
data-structures9
algorithms9
git8
Programming languages (5)
Github contributions (5)
stanfordnlp/stanza
Aug 2019 - May 2022
Stanford NLP Python library for tokenization, sentence segmentation, NER, and parsing of many human languages
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:3 reviews, 271 commits, 18 PRs in 2 years 9 months
Contributions summary:Yuhui made significant contributions to the Stanford NLP Python library, specifically focusing on refactoring the data structure and implementing core components for part-of-speech tagging, lemmatization, and dependency parsing. Their work involved integrating and adapting the CoNLL class, modifying data interfaces, and integrating pre-trained models. They also verified the correctness of the changes by running scripts and comparing results, demonstrating a focus on functionality and testing.
nlppytorchhuman-languagesnamed-entity-recognitionpython
stanford-crfm/helm
Jan 2022 - Dec 2022
Holistic Evaluation of Language Models (HELM), a framework to increase the transparency of language models (https://arxiv.org/abs/2211.09110). This framework is also used to evaluate text-to-image models in HEIM (https://arxiv.org/abs/2311.04287) and vision-language models in VHELM (https://arxiv.org/abs/2410.07112).
Role in this project:ML Engineer
ML EngineerContributions:40 commits, 3 pushes, 3 branches in 10 months
Contributions summary:Yuhui implemented support for multiple commonsense question answering (MCQA) datasets, including HellaSwag, OpenBookQA, CommonSenseQA, PIQA, and SIQA, within the HELM framework. This involved integrating these datasets with the existing multiple-choice question answering pipeline, allowing for evaluation using causal language modeling (CLM) techniques. The user also refactored code for improved clarity and efficiency and corrected computations of calibrated probabilities.
nlparxivabsberthelm
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.