Hi, I’m Zach!

CS PhD, currently working on sparse hierarchical representations for LLMs. Previously, PhD ‘25 focused on Bayesian models + transformers for genomics. Strong generalist with excellent skills in scientific computing + ML.

Contact: hey@zachmaas.com

Previous experience

Role	Timeline	Highlights
Independent Mechanistic Interpretability Researcher	June 2025 - Present	Hierarchical representations for LLMs (research notes) Funded by Coefficient Giving TAIS Grant for 2026
LincSwitch Therapeutics Data Science Consultant	Jun–Dec 2024	Built greenfield data infrastructure (Nextflow, HPC, R) for early-stage biotech
University of Colorado Boulder PhD Researcher	Aug 2020 – May 2025	- BERT + SAEs for genomic regulatory grammar (research notes) - Bayesian MCMC for RNA-seq uncertainty quantification - SOTA cyanobacteria segmentation (U-Net)
University of Colorado Boulder Research Assistant	May 2019 – Aug 2020	Large scale SQL-backed ETL system for genomics pipelines

Stack

Category	Tools
ML/DL	PyTorch, HuggingFace, scikit-learn, PyMC, CUDA, W&B
Data	Pandas, NumPy, SQL, Nextflow, Docker
Languages	Python, Bash/Awk, R, C++, JS/TS, Go, Zig
Infra	Linux, SLURM, Terraform, AWS

Selected Publications

Year	Title	Venue
2025	*Supervised and Unsupervised Methods for Transcriptional Sequencing Data	PhD Thesis
2025	TFIIH kinase CDK7 drives cell proliferation through core TF network	Science Advances
2024	*Internal and External Normalization of Nascent RNA Sequencing	BMC Bioinformatics
2024	*Deconvolution of Nascent Sequencing Data Using TREs	PSB
2021	Transcription Factor Enrichment Analysis	Communications Biology

* = first author

Education

Degree	Institution	Notes
PhD, Computer Science	CU Boulder (2025)	Dissertation: Supervised and Unsupervised Methods for Transcriptional Sequencing Data
B.A., Mathematics & Chemistry	CU Boulder (2019)	Summa Cum Laude