Projects
Experiences
Energy-Efficient AI Direct Researcher
E2S2C Group – Los Angeles, California
- Focusing on efficient training and inference of large models on NLP and computer vision tasks, developed adaptive freezing LoRA as a parameter-efficient fine-tuning (PEFT) method that yields 9.5× fewer trainable parameters on DeBERTaV3 with a SOTA performance in comparison with original LoRA on GLUE benchmark.
- Developed energy-efficient neural networks for hardware implementations targeting tasks such as speech recognition, image classification, and anomaly detection. Pioneered a spiking recurrent model that is efficient for streaming edge computing and has performance comparable to transformer models. In experiments on Google Speech Dataset, our model achieved 50x reduction in number of parameters and 65x reduction in FLOPS in comparison to the SOTA transformer-based models.
- On-going research focusing on long-context compression for LLMs and Linear Transformers.
Machine Learning Engineer
Duarte Lab – La Jolla, California
- Affiliated with CERN, developed variations of graph neural networks (GNNs) combined with generative adversarial networks (GANs) for simulations of high-energy particle collisions using data from the Large Hadron Collider (LHC); Utilized Docker and Kubernetes to ensure efficient deployment and management of the model workflows.
- Engineered Induced set transformers architecture in Message Passing GANs, leading to the conception of the high-performing Induced Generative Adversarial Particle Transformer (iGAPT), realized a linear inference complexity (6× faster) compared to its predecessor and marks high scores based on established evaluation metrics.
Data Scientist Volunteer
CoronaNet – Global
- Collected and updated policies published by Qinghai Province, China in response to COVID-19 to the database.
Education
University of Southern California
08/2023 - 05/2025Master of Science, GPA: 3.91 / 4.0
Major: Electrical and Computer Engineering - Machine Learning and Data Science
- USC-Meta Research and Education in AI (REAL@USC) Center Fellowship, Full Tuition Award
- Ming Hsieh Department of Electrical & Computer Engineering 2024 Outstanding Masters Poster Award
- Masters Students Honors Program
University of California, San Diego
09/2019 - 06/2023Bachelor of Science, GPA: 3.83 / 4.0
Major: Physics; Cognitive Science (Machine Learning & Neural Computation)
Minor: Mathematics
- 2022-2023 Physical Sciences Dean's Undergraduate Award for Excellence (34 selected from 4,000 students)
- Provost Honors
My time
About me
Life is all about learning the latent distribution.
I am curious about everything and love to learn, apply, and share new things.
I use the following technologies:
- Python
- SQL
- C++
- JavaScript
- PyTorch
- Deep Learning
- Data Engineering
- NLP
- Computer Vision
- Fine-tuning
- MongoDB
- Kubernetes
- AWS
- Docker
- Node.js
- React