Gouki Minegishi

I’m a PhD candidate in Technology Management for Innovation at The University of Tokyo, mentored by Professor Yutaka Matsuo.

My research passion lies in mechanistic interpretability, where I unravel the internal mechanism that drive today’s AI systems.

selected publications

  1. ICLR
    SAE.png
    Rethinking Evaluation of Sparse Autoencoders through the Representation of Polysemous Words
    Gouki Minegishi, Hiroki Furuta, Yusuke Iwasawa, and Yutaka Matsuo
    In The Thirteenth International Conference on Learning Representations, ICLR, 2025
  2. ICML
    ICL.gif
    Beyond Induction Heads: In-Context Meta Learning Induces Multi-Phase Circuit Emergence
    Gouki Minegishi, Hiroki Furuta, Shohei Taniguchi, Yusuke Iwasawa, and Yutaka Matsuo
    In Forty-second International Conference on Machine Learning, ICML, 2025
  3. Preprint
    Reasoning_Graph.gif
    Topology of Reasoning: Understanding Large Reasoning Models through Reasoning Graph Properties
    Gouki Minegishi, Hiroki Furuta, Takeshi Kojima, Yusuke Iwasawa, and Yutaka Matsuo
    2025