• line-follower

  A dependency-free TypeScript library for following a stream or append-only file line by line, like tail -F: each complete line arrives as soon as it is written, with its byte offset (in the case of a file). It handles the parts that are easy to get wrong: appends racing the watch setup, coalesced fs.watch events, lines torn across reads, and atomic replacement by rename.

• clauctl

  A CLI for orchestrating Claude agents: humans attach a terminal UI while agents and scripts get the full Claude Agent SDK control surface, including context surgery — rewinding a conversation or reshaping an agent’s effective context down to hand-picked messages plus a summary. See Controlling Claude’s context window.

• renderdag-ts

  My favorite text renderer for DAGs is Meta’s sapling-renderdag. This is a dependency-free TypeScript port of it.

• cargo-ratchet

  A cargo subcommand that keeps resolved dependency versions in Cargo.lock from silently dropping below a committed baseline.

• docket

  A tiny, daemon-less async work queue: a directory you dispatch background commands into and take their captured results back out when they’re ready.

• pictl

  A CLI for orchestrating pi coding-agent instances, letting humans, agents, and scripts drive the same live sessions simultaneously — each on their own terms.

• pi-rpc-rs

  Typed Rust bindings for pi’s RPC mode: runs pi as a subprocess and drives every RPC command from Rust over a correlated event stream.

• gmrfs

  A differentiable version of INLA (Integrated Nested Laplace Approximation) for estimating the parameters of a Gaussian Markov Random Field, so that larger numbers of parameters can be fit with gradient-based methods.

• graph_sampler

  Sequential importance sampling for counting and uniformly sampling the space of molecules, built in service of a machine-learning-for-chemistry project.

• Verily Site Selection Tool

  A differentiable Monte Carlo simulator of a phase 3 clinical trial, built to help a pharma partner plan a COVID vaccine trial: the planner states the forecasts they believe, sees how participant recruitment strategies affect time to success, and can use gradient-based optimization to guide decisions.

Publications

1. Mapping the ionosphere with millions of phones
   Jamie Smith^*, Anton Kast^*, Anton Geraschenko^*, Y. Jade Morton, Michael P. Brenner, Frank van Diggelen, Brian P. Williams
   Nature 635(8038):365–369 (2024). doi:10.1038/s41586-024-08072-x Erratum: Nature 637, E23 (2025).
2. Using phones to map the ionosphere total electron content and improve location accuracy
   Brian Williams, Anton Geraschenko, Anton Kast, Jamie Smith, Jade Morton
   AGU fall meeting abstracts G54A-02 (2023).
3. Integrating deep learning and unbiased automated high-content screening to identify complex disease signatures in human fibroblasts
   Lauren Schiff, Bianca Migliori, Ye Chen, Deidre Carter, Caitlyn Bonilla, Jenna Hall, Minjie Fan, Edmund Tam, Sara Ahadi, Brodie Fischbacher, Anton Geraschenko, Christopher J. Hunter, Subhashini Venugopalan, Sean DesMarteau, Arunachalam Narayanaswamy, Selwyn Jacob, Zan Armstrong, Peter Ferrarotto, Brian Williams, Geoff Buckley-Herd, Jon Hazard, Jordan Goldberg, Marc Coram, Reid Otto, Edward A. Baltz, Laura Andres-Martin, Orion Pritchard, Alyssa Duren-Lubanski, Ameya Daigavane, Kathryn Reggio, NYSCF Global Stem Cell Array Team, Phillip C. Nelson, Michael Frumkin, Susan L. Solomon, Lauren Bauer, Raeka S. Aiyar, Elizabeth Schwarzbach, Scott A. Noggle, Frederick J. Monsma, Daniel Paull, Marc Berndl, Samuel J. Yang, Bjarki Johannesson
   Nature Communications 13:1590 (2022). doi:10.1038/s41467-022-28423-4
4. It's easy to fool yourself: Case studies on identifying bias and confounding in bio-medical datasets
   Subhashini Venugopalan, Arunachalam Narayanaswamy, Samuel Yang, Anton Geraschenko, Scott Lipnick, Nina Makhortova, James Hawrot, Christine Marques, Joao Pereira, Michael Brenner, Lee Rubin, Brian Wainger, Marc Berndl
   NeurIPS 2019 workshop on learning meaningful representations of life (LMRL) (2019). arXiv:1912.07661
5. Applying deep neural network analysis to high-content image-based assays
   Samuel J. Yang, Scott L. Lipnick, Nina R. Makhortova, Subhashini Venugopalan, Minjie Fan, Zan Armstrong, Thorsten M. Schlaeger, Liyong Deng, Wendy K. Chung, Liadan O'Callaghan, Anton Geraschenko, Dosh Whye, Marc Berndl, Jon Hazard, Brian Williams, Arunachalam Narayanaswamy, D. Michael Ando, Philip Nelson, Lee L. Rubin
   SLAS Discovery 24(8):829–841 (2019). doi:10.1177/2472555219857715
6. A “bottom up” characterization of smooth Deligne–Mumford stacks
   Anton Geraschenko^*, Matthew Satriano^*
   International Mathematics Research Notices 2017(21):6469–6483 (2017). doi:10.1093/imrn/rnw201 arXiv:1503.05478
7. Formal GAGA for good moduli spaces
   Anton Geraschenko^*, David Zureick-Brown^*
   Algebraic Geometry 2(2):214–230 (2015). doi:10.14231/AG-2015-010 arXiv:1208.2882
8. Torus quotients as global quotients by finite groups
   Anton Geraschenko^*, Matthew Satriano^*
   Journal of the London Mathematical Society 92(3):736–759 (2015). doi:10.1112/jlms/jdv046 arXiv:1201.4807
9. Toric stacks I: The theory of stacky fans
   Anton Geraschenko^*, Matthew Satriano^*
   Transactions of the American Mathematical Society 367(2):1033–1071 (2015). doi:10.1090/S0002-9947-2014-06063-7 arXiv:1107.1906
10. Toric stacks II: Intrinsic characterization of toric stacks
    Anton Geraschenko^*, Matthew Satriano^*
    Transactions of the American Mathematical Society 367(2):1073–1094 (2015). doi:10.1090/S0002-9947-2014-06064-9 arXiv:1107.1907
11. There is no degree map for 0-cycles on Artin stacks
    Dan Edidin^*, Anton Geraschenko^*, Matthew Satriano^*
    Transformation Groups 18(2):385–389 (2013). doi:10.1007/s00031-013-9225-x arXiv:1208.3239
12. Toric stacks
    Anton Geraschenko
    PhD thesis, University of California, Berkeley (2011). Adviser: Vera Serganova.

^* co-first author. In mathematics, author order is alphabetical. See also my Google Scholar profile.