AI People Wiki

Builders & open source

Sam Gross

Company listed on Sam Gross's GitHub profile, Thinking Machines Lab

Sam Gross is the second author of the 2019 PyTorch paper, listed there with Facebook AI Research, and PyTorch's maintainers page lists Gross as an emeritus maintainer of torch.nn and of the MKLDNN integration. Gross is the author of PEP 703, "Making the Global Interpreter Lock Optional in CPython," whose status is Final. Gross's GitHub profile lists Thinking Machines as their company.12345

At a glance

Current role
Company listed on Sam Gross's GitHub profile
Sector
Company
Work listed
2
Sources cited
11
Profile reviewed

Research

Gross co-authored papers across vision, language and games. These include a study of learning physical intuition from block towers (2016), the MultiPath network for object detection (2016), "Hard Mixtures of Experts for Large Scale Weakly Supervised Vision" (2017, as first author), Deep Counterfactual Regret Minimization (2018) and the fairseq sequence-modelling toolkit (2019).678910

Open-source work

Gross is the second author of the 2019 PyTorch paper, which describes the library's imperative, Pythonic design and its implementation, and is listed on PyTorch's maintainers page as an emeritus maintainer of torch.nn and of the MKLDNN integration.123

Gross wrote PEP 703, created in January 2023, which proposed a build configuration (--disable-gil) that lets CPython run Python code without the global interpreter lock, along with the changes needed to make the interpreter thread-safe. The PEP targets Python 3.13, was resolved in October 2023 and now has Final status.4

Work and contributions

  • PyTorch Open-source code

    Second author of "PyTorch: An Imperative Style, High-Performance Deep Learning Library" (2019); listed on PyTorch's maintainers page as an emeritus maintainer of torch.nn and MKLDNN. The repository is under a BSD-style license.1311

  • PEP 703: Making the Global Interpreter Lock Optional in CPython Published research

    Author of PEP 703 (created January 2023, status Final), the Python Enhancement Proposal for making the global interpreter lock optional in CPython.4

Timeline

  1. Co-authored "Learning Physical Intuition of Block Towers by Example" with Adam Lerer and Rob Fergus.6
  2. Co-authored "A MultiPath Network for Object Detection".7
  3. Posted "Hard Mixtures of Experts for Large Scale Weakly Supervised Vision" as first author.8
  4. Co-authored "Deep Counterfactual Regret Minimization".9
  5. Co-authored fairseq, a toolkit for sequence modeling.10
  6. Posted the PyTorch paper as second author.1
  7. Created PEP 703, "Making the Global Interpreter Lock Optional in CPython".4
  8. PEP 703 was resolved, for the Python 3.13 release series.4

Other documented roles

Roles the sources document that are not confirmed as current.

Where they work

Suggest a correction or removal

Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.

References

11 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026

  1. PyTorch: An Imperative Style, High-Performance Deep Learning Library (arXiv 1912.01703)arXiv · Paper · published Dec 3, 2019 · accessed Oct 21, 2026
  2. PyTorch: An Imperative Style, High-Performance Deep Learning Library, paper PDF with author affiliations (arXiv 1912.01703)arXiv · Paper · published Dec 3, 2019 · accessed Oct 21, 2026
  3. PyTorch Governance | Maintainers (PyTorch 2.14 documentation)PyTorch · Documentation · published Jun 5, 2026 · accessed Oct 21, 2026
  4. PEP 703 – Making the Global Interpreter Lock Optional in CPythonPython Software Foundation (peps.python.org) · Documentation · published Jan 9, 2023 · accessed Oct 21, 2026
  5. colesbury (Sam Gross) on GitHubGitHub · Official page · accessed Oct 21, 2026
  6. Learning Physical Intuition of Block Towers by Example (arXiv 1603.01312)arXiv · Paper · published Mar 3, 2016 · accessed Oct 21, 2026
  7. A MultiPath Network for Object Detection (arXiv 1604.02135)arXiv · Paper · published Apr 7, 2016 · accessed Oct 21, 2026
  8. Hard Mixtures of Experts for Large Scale Weakly Supervised Vision (arXiv 1704.06363)arXiv · Paper · published Apr 20, 2017 · accessed Oct 21, 2026
  9. Deep Counterfactual Regret Minimization (arXiv 1811.00164)arXiv · Paper · published Nov 1, 2018 · accessed Oct 21, 2026
  10. fairseq: A Fast, Extensible Toolkit for Sequence Modeling (arXiv 1904.01038)arXiv · Paper · published Apr 1, 2019 · accessed Oct 21, 2026
  11. pytorch/pytorch LICENSE (BSD-style license)PyTorch (GitHub repository) · License · accessed Oct 21, 2026