Home
Miscellaneous

Sophia Lu

 

Research Interests

Efficient sampling algorithms for black-box models, generative modeling and synthetic data, distributional regression, probabilistic modeling and Bayesian methods, approximate Bayesian inference, and their applications to Bioengineering, Genetics and Genomics, Astrophysics, Econometrics, etc.

About

I am a final-year Ph.D. candidate in the Department of Statistics at Stanford University, where I am fortunate to be advised by Wing H. Wong.

I develop statistical methods for modeling, inference, and sampling in complex probabilistic and black-box systems, drawing on modern deep-learning and generative models. My research spans Bayesian modeling and scalable inference, sampling for pretrained foundation models, LLM-assisted inference, generative and resampling methods, and synthetic data generation and distributional learning, with applications in bioengineering, genomics, astrophysics, and econometrics.

Before pursuing my doctoral studies, I graduated with honors in Mathematical and Computational Science and a minor in Mathematics from Stanford University. My research has been supported by the Stanford Data Science Graduate Fellowship, the Two Sigma Graduate Fellowship Fund, the Google Cloud Academic Research Grant (Co-PI), and the Stanford HAI-Google Cloud Credits Grant (Co-PI).

Research

Patents

ProVADA protein-redesign framework
Patent pending
U.S. provisional patent application filed in 2026.

Publications and Preprints

  1. Generative steering to adapt proteins beyond native biological contexts
    Xiaowei Zhang*, Wenhui Sophia Lu*, Ben Viggiano*, Kevin Reynolds, Sophia Burick, Elena Sotillo, Haoyu Dai, Geonhu Lee, Zixuan Liu, Nuoya Yang, Luis S. Mille-Fragoso, Euan Ashley, Bingfei Yu, Crystal Mackall, Wing Hung Wong, Xiaojing Gao
    Submitted

  2. FinMFT: Multi-Source Fine-Tuning of Language Models for Financial News
    Silu Li, Wenhui Sophia Lu, John Mulvey, Siyu Li
    Accepted for publication at the 7th ACM International Conference on AI in Finance (ICAIF 2026)

  3. CuGen: A GPU-Accelerated Framework for Large-Scale Genomics
    Tuomo Kiiskinen, Joshua Richland, William Wang, Wenhui Sophia Lu, Balasubramanian Narasimhan, Trevor Hastie, Robert Tibshirani, Manuel A. Rivas
    Submitted to Nature Genetics

  4. Fidelity-Concordance Reward Calibration for Reliable Generative Design
    Wenhui Sophia Lu, Xiaowei Zhang, Xiaojing J. Gao, Dominik Rothenhaeusler, Wing Hung Wong
    ICML 2026 Workshop on Generative and Agentic AI for Biology

  5. BUDDY: Blending Training and Deployment Data with Weighted Expert Ensembles for Post-Hoc LLM Calibration
    Aishwarya Mandyam*, Wenhui Sophia Lu*, Wing Hung Wong, John Duchi, Barbara E. Engelhardt
    ICLR 2026 Workshop on Principled Design for Trustworthy AI: Interpretability, Robustness, and Safety across Modalities

  6. Univariate-Guided Sparse Regression for Biobank-Scale High-Dimensional Omics Data
    Joshua Richland, Tuomo Kiiskinen, William Wang, Wenhui Sophia Lu, Balasubramanian Narasimhan, Trevor Hastie, Manuel Rivas, Robert Tibshirani
    PLOS Genetics 22(9): e1012314 (2026)

  7. Steering Protein Generative Models at Test Time for Guided AAV2 Capsid Design
    Ben Viggiano*, Wenhui Sophia Lu*, Xiaowei Zhang*, Luis Santiago Mille-Fragoso, Xiaojing J. Gao, Euan Ashley, Wing Hung Wong
    Accepted for publication and selected for an oral presentation at the Pacific Symposium on Biocomputing (PSB) 2026

  8. ProVADA: Generating Subcellular Protein Variants via Ensemble-Guided Test-Time Steering
    Wenhui Sophia Lu*, Xiaowei Zhang*, Luis Santiago Mille-Fragoso, Haoyu Dai, Xiaojing J. Gao, Wing Hung Wong
    Spotlight & Oral at Generative AI for Biology Workshop, ICML

  9. Likelihood-Free Adaptive Bayesian Inference via Nonparametric Distribution Matching
    Wenhui Sophia Lu, Wing Hung Wong

  10. Generative Modeling for Tabular Data via Penalized Optimal Transport Network
    Wenhui Sophia Lu*, Chenyang Zhong*, Wing Hung Wong
    Package available here
    Under revision

  11. Comparison of REML methods for the study of phenome-wide genetic variation
    Damian Pavlyshyn, Wenhui Sophia Lu, Iain M. Johnstone, and Jacqueline L. Sztepanacz
    Under revision at Genetics

  12. Sc-compReg enables the comparison of gene regulatory networks between conditions using single-cell data
    Zhana Duren*, Wenhui Sophia Lu*, Joseph G. Arthur, Preyas Shah, Jingxue Xin, Francesca Meschi, Miranda Lin Li, Corey M. Nemec, Yifeng Yin, and Wing Hung Wong
    Nature Communications, 2021

* indicates equal contribution

Invited and Contributed Talks

  • Likelihood-Free Adaptive Bayesian Inference via Nonparametric Distribution Matching

    • 2026 ICSA Applied Statistics Symposium, Arlington, Virginia, Jun. 2026.

    • Gratefully selected for ICSA Student Travel Award.

  • Likelihood-Free Adaptive Bayesian Inference via Nonparametric Distribution Matching

  • Efficient Likelihood-Free Adaptive Bayesian Inference

  • Efficient Generative Modeling via Penalized Optimal Transport Network

  • ProVADA: Generation of Subcellular Protein Variants via Ensemble-Guided Test-Time Steering

  • Likelihood-Free Adaptive Bayesian Inference

  • An Introduction to Approximate Bayesian Computation and Likelihood-Free Inference

    • Guest lectures (3 sessions), Applied Bayesian Statistics | STATS 371, Stanford University, May 2025.

  • Towards faithful synthetic data generation via penalized optimal transport network

  • Modern Bayesian Modeling and Adaptive Bayesian Inference

    • Guest lecture, Topics in Computing for Data Science | STATS/BIODS 352, Stanford University, Apr 2025.

  • Towards faithful synthetic data generation via penalized optimal transport network

  • Towards faithful synthetic data generation via penalized optimal transport network

    • Lightning talk, Women in Data Science Workshop, Stanford University, Mar 2025.

  • Modern Bayesian modeling and adaptive Bayesian inference

    • Citadel GQS PhD Colloquium, New York, May 2024.

Selected Presentations

  • Efficient Likelihood-Free Adaptive Bayesian Inference

    • Stanford Berkeley Joint Colloquium, Stanford University, October 2025.

  • ProVADA: Generation of Subcellular Protein Variants via Ensemble-Guided Test-Time Steering

    • Stanford Bio-X Interdisciplinary Initiatives Seed Grants Poster Session, Aug 2025.

    • Selected as one of the 11 Rank 1 poster awards (among 336 poster submissions).

  • Efficient Generative Modeling via Penalized Optimal Transport Network

    • Berkeley Stanford Joint Colloquium, UC Berkeley, May 2025.

  • Efficient Generative Modeling via Penalized Optimal Transport Network

    • The Past, Present & Future of Statistics in the Era of AI, The George Washington University, May 2025.

    • Gratefully supported by travel award from NSF.

  • Efficient Generative Modeling via Penalized Optimal Transport Network

    • Optimization and Statistical Learning Workshop, Columbia University, Apr 2025.

    • Gratefully supported by travel award from NSF.

  • Efficient Generative Modeling via Penalized Optimal Transport Network

    • Statistics and Optimal Transport Workshop, Columbia University, Mar 2025.

    • Gratefully supported by travel award.

  • Efficient Generative Modeling via Penalized Optimal Transport Network

    • Stanford Data Science Conference, Stanford University, Apr 2024.

Contact

sophialu (at) stanford (dot) edu