Profile picture

하주헌 JuHeon

Stories only I can tell

M.S. student (Ajou Univ., Artificial Intelligence)

  • LLM Interpretability
  • LLM Safety & Alignment
  • Representation Engineering
  • Activation Steering
  • Human-Centered AI

Publications

EMNLP 2026 Main ConferenceFirst Author

Abstract

This work investigates the activation-level structure of supportive empathy in instruction-tuned LLMs. Using the EPITOME framework and contrastive activation steering, we examine the controllability and representational geometry of emotional reactions, interpretations, and explorations, as well as their interaction with persona-induced activation shifts.

  • All (46)
  • Essay (17)
  • etc (2)
  • Python (1)
  • React (6)
  • JS/TS (3)
  • Vue (16)
  • HTML/CSS (1)
  • All (46)
  • Essay (17)
  • etc (2)
  • Python (1)
  • React (6)
  • JS/TS (3)
  • Vue (16)
  • HTML/CSS (1)