Ceyhun Efe Kayan

Ph.D. Student, Computer Science · Drexel University

I am a second-year Ph.D. student in Computer Science at Drexel University, advised by Li "Harry" Zhang. I study the mechanistic interpretability of large language models: how LLMs internally represent and carry out reasoning, planning, and formalization. I use activation steering, representation geometry, and circuit-level analysis to identify the mechanisms behind these behaviors and to intervene on them.

In 2026 I was a Research Scientist Intern at IBM Silicon Valley Lab, working on applications of LLM hidden states. Before my Ph.D., I received an M.S. in Machine Learning Engineering from Drexel and a B.S. in Electrical and Electronics Engineering from Izmir Institute of Technology.

[email protected] Google Scholar GitHub LinkedIn X CV (PDF)

News

• Sep 2026 Prototype-Based Dynamic Steering for Large Language Models accepted to the EMNLP 2026 Main Conference.
• Aug 2026 A Reality Check of Language Models as Formalizers on Constraint Satisfaction Problems selected as a spotlight at the COLM 2026 Workshop on Efficient Reasoning.
• Jun 2026 Started a Research Scientist Internship at IBM Silicon Valley Lab, San Jose.
• Sep 2025 Started my Ph.D. in Computer Science at Drexel University.

Selected Publications

• Prototype-Based Dynamic Steering for Large Language Models
  Ceyhun Efe Kayan, Li Zhang
  EMNLP 2026 (Main Conference), 2026
  arXiv
• A Reality Check of Language Models as Formalizers on Constraint Satisfaction Problems
  Rikhil Amonkar^*, Ceyhun Efe Kayan^*, Qimei Lai, Ronan Le Bras, Li Zhang
  COLM 2026 Workshop on Efficient Reasoning, 2026 Spotlight
  arXiv
• Intensity and Phase Stacked Analysis of a Φ-OTDR System Using Deep Transfer Learning and Recurrent Neural Networks
  Ceyhun Efe Kayan, Kivilcim Yuksel Aldogan, Abdurrahman Gumus
  Applied Optics 62(7), 2023
  Paper DOI

* Equal contribution. See all publications.