TY  - RPRT
TI  - Neologism Learning as a Parameter-Efficient Alternative to Fine-Tuning for Model Steering
AU  - Sungjoon Park
AU  - Varun Ramamurthi
AU  - Owen Terry
PY  - 2025
UR  - https://arxiv.org/abs/2512.18551
ID  - 2512.18551
ER  -