Qijia He

I am a PhD student in the Department of Statistics at the University of Washington, advised by Prof. Bo Zhang. I collaborate with Prof. Ranjay Krishna at the UW CSE RAIVN Lab on multimodal AI research. Before graduate school, I received my B.S. in Statistics from Sun Yat-sen University. I also have industry experience as an intern at Apple and TikTok.

My research interests lie broadly in vision–language models and causal machine learning, with a focus on multimodal reasoning and generation, agentic AI systems, and causal machine learning.

Contact me: heqj3@uw.edu

Selected Research

LLM / VLM

VFIG is the first large-scale ecosystem for scientific figure-to-SVG generation, introducing a 66K figure–SVG dataset, a new benchmark, and a family of VLMs trained with SFT+RL that achieve state-of-the-art open-source performance and parity with GPT-5.2.

Large language models (LLMs) generate rationales that improve reasoning and interpretability, but existing evaluations using binary human or LLM preferences are limited and opaque. We introduce an attribute-based evaluation framework that defines key rationale qualities, explains human preferences, and enables more nuanced model comparisons through attribute-specific analysis.

Causal Inference

We proposed a framework for using placebo samples in observational studies to detect and correct for unmeasured confounding bias. It develops identification assumptions and estimation methods—including regression, weighting, and doubly robust approaches—and validates them through simulations and an applied case study on tax credits and infant health.

We developed a potential outcomes framework to estimate the ITT effect of an active control versus placebo in active-controlled trials, using historical placebo-controlled data. Our method enables ITT estimation when the placebo arm is unavailable and accounts for unmeasured confounders using instrumental variables.

We introduced a new criterion for individualized treatment rules based on the adjusted probability of longer survival, offering a clear and clinically relevant objective. Our method, optimal adjusted probability learning, constructs the best treatment rule by maximizing this nonparametric survival benefit.

Industry Experience

E-commerce Governance Algorithms, TikTok

Ads Platform, Apple

Teaching Experience

• STAT 311 Elements of Statistical Methods (Winter 2024, Autumn 2025)
• STAT 513 Statistical Inference (Winter 2026)
• STAT 435 Introduction to Statistical Machine Learning (Spring 2026)

Tutor in Mathematical analysis (Fall 2018)

Teaching Assistant in primary-school Olympiad Mathematics (2017-2018)