Ethan Yip
Principal Engineer at Resonance | Researcher at Substrate Labs
Research engineer focused on mechanistic interpretability and applied ML in safety and global systems.
Currently, I'm an Engineer at Resonance working on the initial Humanoid Architectures and Voice Audio systems. In addition, I research at Substrate Labs on interpretability and applied machine learning across agents, physics, and alignment.
I enjoy designing and analyzing systems while pursuing knowledge in ML, EECS, Philosophy, and societal structures.
When I'm not pursuing work in the above, I like to read into classical latin, ancient history and psychology!
Research
PNS Guard: Geometric AI Safety and Active Defense Framework
Applied research establishing a geometric 3D security "MRI" framework using Principal Nested Sphere (PNS) coordinates to isolate and block transformer activation jailbreaks.
[+]Before Measuring Human-Agent Teams: Validating Human References in Agent Benchmarks
Audits the reliability of human task-difficulty baselines behind agent benchmarks like METR, SWE-bench Verified, and GAIA, proposing a validity protocol for claims that agents can substitute for human work.
The Survey Item Predicts Human Disagreement Better Than Simulator Output Entropy
Demonstrates that a small text encoder reading only a survey question predicts human disagreement better than an LLM social simulator's own output uncertainty.
Experience
Principal Engineer
May 2026 – Present · San Francisco Bay Area (Hybrid)
Youth Delegate & Director of HTSI
Sep 2025 – Present · New York City Area (Part-time)
Venture Strategy & Ops
2025 · San Francisco Bay Area (Remote)
Awards & Recognition
John Locke Essay Competition — "Very High Commendation"
Aug 2025
ACL/NJCL National Latin Exam — 2x Gold, Summa Cum Laude (Perfect Score)
May 2026
European Challenge — 2x National Semifinalist
Apr 2025