PeopleofSuperintelligence

Researcher

Yingzi Ma

Also known as 马英梓

Yingzi Ma is a PhD student in Information Science at the University of Wisconsin–Madison who works on language models, autonomous driving, and model safety. They studied Tourism Management and earned an Artificial Intelligence B.Eng. at Sichuan University, and began research in Shanghai Jiao Tong University's XLANCE lab. Their papers include MixLoRA for efficient fine-tuning, Dolphins for conversational driving, FIUBench for vision-language-model unlearning, and ARMOR for language-model safety. In June 2026, they joined NVIDIA's Autonomous Vehicle Research group as a summer research intern.

Badges

Timeline

  1. 2019–2024

    Sichuan University and XLANCE

    At Sichuan University, they study Tourism Management from 2019 to 2020 and earn a B.Eng. in Artificial Intelligence there in 2024. In April 2022, they join Shanghai Jiao Tong University's XLANCE lab as an intern in its language-model group, advised by Lu Chen.

  2. 2024

    MixLoRA for language models

    In 2024, they coauthor MixLoRA, a parameter-efficient fine-tuning method that places LoRA experts inside a frozen language model and uses a top-k router.

  3. 2023–2024

    Dolphins for autonomous driving

    In December 2023, they coauthor Dolphins, a vision-language model for conversational driving assistance that processes video, text instructions, and past control signals. The paper is accepted to ECCV 2024.

  4. 2024–2025

    Benchmarking VLM unlearning

    In 2024, they coauthor FIUBench, a benchmark for evaluating whether vision-language models can forget fictitious identities under a right-to-be-forgotten setting. The work is accepted to ICLR 2025.

  5. 2025–2026

    ARMOR for language-model safety

    In 2025, they coauthor ARMOR, which analyzes jailbreak strategies, extracts user intent, and applies policy-based safety verification. The paper is accepted to ICLR 2026.

  6. 2025–2026

    dVLM-AD for end-to-end driving

    In December 2025, they coauthor dVLM-AD, a diffusion-based vision-language model that unifies perception, structured reasoning, and low-level planning for end-to-end driving. The paper is accepted to CVPR 2026 Findings.

  7. June 2026

    Summer research at NVIDIA

    In June 2026, they join NVIDIA's Autonomous Vehicle Research group as a research intern for the summer.

  8. As of October 2026

    PhD student at UW–Madison

    As of October 2026, they are a second-year PhD student in Information Science at the University of Wisconsin–Madison, advised by Chaowei Xiao.

Career

  1. 2026NVIDIAResearch intern
  2. From 2022Shanghai Jiao Tong UniversityIntern

Education

  1. 2024Sichuan UniversityB.Eng. in Artificial Intelligence
  2. 2020Sichuan UniversityTourism Management
  3. University of Wisconsin–MadisonPhD

From the biography.

  1. 16% of AI career moves in the 2020s went from a university to a companyIncludes Yingzi Ma. We estimate that 16% of AI career moves in the 2020s went from a university to a company; 29% of moves starting at a university went to a company, and 76% ended the university role.
  2. 84% of people shaping AI with recorded undergraduate and doctoral schools changed institutions for their doctorateIncludes Shanghai Jiao Tong University. 84% of people shaping AI with recorded undergraduate and doctoral schools changed institutions for their doctorate. Across 2,116 people with both an undergraduate degree and a doctorate recorded, we trace which undergraduate schools feed which doctoral programs, how often students stay at the same school, and how paths differ by birth-country region.
  3. Google has the highest combined h-index among AI organizationsIncludes NVIDIA. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.
  4. 20% of people at companies who shape AI also hold a university role, led by Google and Stanford UniversityIncludes NVIDIA. We estimate that 539 of the 6,766 people with a known current organization (8%) hold both a university role and a company role, including 20% of people at companies and 13% of people at universities.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Yingzi Ma yet. Send us a free-licensed photo.

  1. Yingzi Ma gray311.github.io. Checked 20 October 2026.
  2. MixLoRA: Enhancing Large Language Models Fine-Tuning with LoRA-based Mixture of Experts arxiv.org. Checked 20 October 2026.
  3. Dolphins: Multimodal Language Model for Driving research.nvidia.com. Checked 20 October 2026.
  4. Benchmarking Vision Language Model Unlearning via Fictitious Facial Identity Dataset arxiv.org. Checked 20 October 2026.
  5. ARMOR: Aligning Secure and Safe Large Language Models via Meticulous Reasoning arxiv.org. Checked 20 October 2026.
  6. dVLM-AD: Enhance Diffusion Vision-Language-Model for Driving via Controllable Reasoning arxiv.org. Checked 20 October 2026.