AS
Amon Suzuki
Developer / Researcher
📍 Based in Tokyo · Remote friendly
Email GitHub LinkedIn
Open for collaborations & new roles

Hi, I'm Amon Suzuki

I'm a researcher focused on Mechanistic Interpretability and AI4Science, working to fill the gap between Human and AI research capability for a coworking future.

Mechanistic Interpretability AI4Science
Experience
2+ yr
Recent focus
ASIC/ML
Time zone
GMT+9
Currently accepting selected freelance / side projects.
View selected work → Contact
About

A short introduction

A researcher focused on Mechanistic Interpretability and AI4Science to bridge the gap between Human and AI research capability.

I'm a researcher focused on Mechanistic Interpretability and AI4Science. My goal is to fill the gap between Human and AI research capability, enabling a future where humans and AI cowork effectively on scientific problems.

I focus on understanding how neural networks work internally (mechanistic interpretability) and applying AI to scientific discovery (AI4Science) to bridge the gap between what humans and AI can each contribute to research. I also work on Spiking Neural Networks on ASIC and ROS/Autoware tools for autonomous vehicles.

I'm currently exploring opportunities in mechanistic interpretability research and AI4Science, with a passion for building a coworking future between humans and AI.

Biography

経歴 / Achievements

主な活動・受賞歴

2024年3月
42 Tokyo
合格
2025年7月
NICT Quantum Camp 探索型人材育成コース
採択
2025年8月
セキュリティ・キャンプ2025全国大会 L2「高位言語から物理レイアウトまで半導体設計ワークフローを経験して MyLSI を発注しよう！」ゼミ
修了
2025年9月
自動運転AIチャレンジ予選
総合1位通過
2025年10月
自動運転AIチャレンジ決勝
総合優勝　日本自動車工業会会長賞
Work

Selected projects

A selection of my recent projects showcasing hardware design, machine learning, and software development.

2025 · Personal
Reward-implemented Spiking Neural Network
SNN that learns through STDP and Reward Feedback, designed for ASIC implementation with integrated on-chip memory using Tiny Tapeout.
Verilog · ASIC · Neural Networks
Role: Hardware Design & Research View
2025 · Personal
CosenseGraph
A 3D visualization tool for Cosense (formerly Scrapbox) that displays page relationships as an interactive node graph with LLM-powered clustering.
TypeScript · OpenAI API · 3D Visualization
Role: Full Stack Development View
2025 · Tools
MCAP Reader for Autoware
A ROS bag analyzer adapted for MCAP files, enabling trajectory analysis and sensor data visualization for autonomous vehicles.
Python · ROS · Autoware
Role: Development & Analysis View
Contact

Say hello

Whether you're interested in working together, have a question, or just want to share an idea, feel free to get in touch.

The fastest way to reach me is by email, but I'm also occasionally active on GitHub and LinkedIn. Short introductions are welcome.

If you already have a project in mind, it helps to include your timeline, rough scope, and any links to references or existing work.

Preferred contact hours: weekday evenings (GMT+9), but I'll reply as soon as I can.
amonsuzuki@keio.jp GitHub LinkedIn Resume / CV
This form is only a visual example. Connect it to your backend or a service like Formspree / Netlify Forms.
No spam. I'll only use your details to reply.