Kyle R. Chickering - krchickering@ucdavis.edu
Updated: 05/24/2026
UC Davis - 2025. Advisor: Prof. Muhao Chen.
Thesis: Self-Similar Shock Formation and Some Results in Deep Learning
UC Davis - 2021. Advisor: Prof. Steve Shkoller.
UC Davis - 2019. Advisor: Prof. Timothy J. Lewis.
Zhaonan Li, Kyle R. Chickering, Bangzheng Li, Jacob Dineen, Xiao Ye, Zhikun Xu, Shijie Lu, Yuxi Huang, Ming Shen, Bach Nguyen, Jaya Adithya Pavuluri, Mau Son Nguyen, Sanika Chavan, Ngoc Minh Thu Le, Muhao Chen, and Ben Zhou. VisAnalog: A Diagnostic Suite for Visual Concept Transfer on Natural Images. (2026). [arXiv:2605.23141]. Workshop on Visual Concepts at CVPR 2026.
Kyle R. Chickering, Huijuan Wang, Mengxi Wu, Alexander Moreno, Muhao Chen, Xuezhe Ma, Daria Soboleva, Joel Hestness, Zhengzhong Liu, and Eric Xing. GQA-μP: The maximal update parameterization for grouped query attention. (2026). [arXiv:2605.15290]. Submitted.
Kyle R. Chickering, Bangzheng Li, and Muhao Chen. QLIP: A Dynamic Vision Prior Enhances MLLM Performance Without Retraining. ICLR 2026. [arXiv:2505.23004].
Kyle R. Chickering. A Quasilinear Algorithm for Computing Higher-Order Derivatives of Deep Feed-Forward Neural Networks. (2024). [arXiv:2412.09752].
Kyle R. Chickering, Ryan C. Moreno-Vasquez, and Gavin Pandya. Asymptotically self-similar shock formation for the 1d fractal Burgers equation. Siam J. Math. Anal. Vol 55, Iss. 6 (2023). [PDF, 2021].
November 2025 - Present
May 2025 - October 2025
September 2019 - September 2025
Published a pure-mathematics research paper in a top journal my second year of graduate school by leading a group of graduate students studying shock-formation for nonlocal fluid models.
Implemented a SotA deep-learning model for solving differential equations which outperforms all reproducible SotA models in the literature by a factor of 1000x by leveraging my knowledge of deep-learning fundamentals to well-condition the training Hessian (paper in progress).
Improved model training times by a factor of 3x by adapting novel research algorithms for use with my models.
Founded and led the Math Directed Reading Program, which pairs curious undergraduates with graduate student mentors for small-group directed learning. Served as Treasurer of the graduate group for two years.
June 2023 - October 2024
Reduced prototype foundation video model training time by a factor of 7.5x by implementing a novel recursive unrolling algorithm.
Expanded temporally-consistent video generation length from 3.5 seconds to over 10 seconds by building a Google VIDIM-inspired frame prediction model.
Cut SSR model memory footprint by 40% by transitioning to SotA architecture.
Increased GPU utilization from 40% to 80% through profiling, distributed training improvements, and more efficient data-loading.
Trained Irreverent’s first viable foundation video model by developing a cluster management framework to do distributed training over our 64-node cluster.
October 2023 - June 2024
Responsible for building and scaling the engineering and research infrastructure including choosing Python, PyTorch, and C++ as our tech stack, and implementing software engineering best practices from day zero.
Accelerated prototype algorithm performance by over 100x by re-writing critical computational code in C++.
Saved investors over 100k by making the strategic decision to abandon our efforts after conducting extensive research and experimentation and concluding that we could not bring our product to market in the desired timeline.
May 2018 - September 2019
May 2016 - September 2018
Acquired new customers by working with the technical marketing team to create, launch, and deploy a network provisioning plugin. [GitHub].
Brought the company’s outdated Ansible codebase into compliance by refactoring thousands of lines of Python code.
Built and deployed an Android application to do real-time facial recognition and tracking by using existing computer-vision libraries. [GitHub1], [GitHub2].
$n$-TangentProp: A Quasilinear Algorithm for Computing Derivatives of Neural Networks
UC Davis Student Run Analysis and PDE Seminar. January 2025.
Asymptotic Self-Similar Shock Formation for 1D Fractal Burgers
UC Davis Student Run Analysis and PDE Seminar. April 2021.
Long Time Existence via a Modified Energy Method
UC Davis Student Run Analysis and PDE Seminar. February 2020.
Effect of Elastic Fluids on Cilia Dynamics
UC Davis Undergraduate Mathematics Research Conference. October 2019.
Measure Theory, Integration, and $L^p$ Spaces
Undergraduate Math and Physics Reading Group. April 2019.
Mathematical Modeling of Anomalous Electrical Activity in Networks of TSC1- Mutant Neurons in the Thalamus
UC Davis Spring Research Conference. April 2019.
Fractals and Recursion
UC Davis Math Circle. March 2019.
LaTeX: A Document Preparation System
UC Davis Math Club. February 2019.
Spike-Escape Behavior in Electrically Coupled Thalamocortical Relay Cells with TSC1 Gene Deletion
UC Davis Undergraduate Mathematics Conference. October 2018.
Mathematics Department Fellowship
Awarded by the Mathematics department. Fall 2020.
Citation for Excellence in Undergraduate Research
Awarded by the Mathematics department. October 2019.
Mathematics Department Summer Fellowship
Summer 2019.
Citation for Outstanding Performance
June 2019.
UC Davis Outstanding Senior Award in Applied Mathematics
Awarded by the university to a single student each year. May 2019.
2025 - Present
ICLR
CVPR
2021
Co-founded with Ashleigh Adams.
2020 - 2021