Qiankang (Kant) Wang
Machine learning for biology — representation learning, foundation models, and scientific computing.
qkwang@berkeley.edu ·
GitHub ·
LinkedIn
Education
B.A. in Data Science (Data Science Honors Program), University of California, Berkeley — expected 2027. GPA 3.82 / 4.00. Dean's List 2025.
Curriculum vitae · PDF
Research
- Stanford University — Undergraduate Research Assistant, 2026–Present. Cross-model interpretability of genomic sequence models (AlphaGenome, ChromBPNet).
- Computational Biophysics Lab, UC Irvine — Undergraduate Research Assistant, 2024–2026. GPU-accelerated Poisson–Boltzmann solvers in AmberTorchPB; Slurm pipeline for over one million PBSA calculations.
Publication
Wu, Y., Wang, Q., Jiang, R., Luo, R. (2026).
AmberTorchPB: A Unified Framework for Poisson–Boltzmann-Based Reaction Field Energy Calculation via Tensor Computation.
Journal of Chemical Theory and Computation.
Selected projects
- simclr-pytorch — SimCLR-style self-supervised learning on Tiny-ImageNet-200 with frozen-feature k-NN evaluation.
- personagent — self-evolving persona LLM agent with an evaluation-to-feedback loop, deployed via a platform-neutral gateway.
- Decision-Tree — from-scratch C++ decision-tree classifier (Gini impurity, Titanic dataset).
Skills
ML: PyTorch, LibTorch, TensorFlow, transformers, self-supervised learning, diffusion models.
Interpretability: DeepLIFT / DeepSHAP, TF-MoDISco, FiNeMo, in silico perturbation.
Scientific computing: GPU optimization, iterative solvers (CG, BiCG, GMRES, SOR), Poisson–Boltzmann / PBSA, molecular simulation.
Languages: Python, C++, Java, MATLAB, Bash, SQL.
Tools: Linux, Git, Docker, CMake, Slurm, Jupyter, LaTeX.
This is a text-only fallback. The interactive site requires JavaScript.