Curt Tigges
Member Of Technical Staff at Goodfire
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolCurt Tigges is a mechanistic interpretability researcher and Member of Technical Staff based in San Francisco with eight years of experience translating deep model behaviors into usable tools and libraries. He has led engineering and science efforts at Decode Research and EleutherAI, building widely used open-source toolkits (Probity, SAELens, CLT training pipelines) and large-scale circuit analysis pipelines spanning models from 70M to 13B parameters. Curt combines hands-on ML engineering—multi-GPU training, activation/path-patching tooling, and production activation pipelines—with applied data science and optimization experience from prior industry roles. His work bridges rigorous research and practical infrastructure, shipping reproducible analysis platforms like Neuronpedia while mentoring teams on mechanistic techniques. An MCS (Data Science) from UIUC, he’s notable for operationalizing cutting-edge mech interp methods into public libraries that other researchers rely on.
8 years of coding experience
6 years of employment as a software developer
Bachelor of Science (B.S.) Science Technology and Society, Bachelor of Science (B.S.) Science Technology and Society at North Carolina State UniversityMaster of Computer Science Data Science track, Master of Computer Science Data Science track at University of Illinois Urbana-Champaign
English, Japanese, Chinese
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (11)
huggingface-transformers10
deepspeed10
machine-learning9
gpt9
parallel9
transformers9
pytorch9
deep-learning8
language-model8
encoding2
tensorflow1
Programming languages (3)
Github contributions (5)
curt-tigges/circuits-over-time
Mar 2023 - May 2024
Contributions:23 PRs, 295 pushes, 4 branches in 1 year 2 months
jbloomAus/SAEDashboard
Aug 2024 - Feb 2025
Contributions:28 PRs, 53 pushes, 17 branches in 5 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.