On Mitigation of Subliminal Learning in Large Language Models
Liminal training is introduced, an annealed KL-regularized fine-tuning method that constrains early drift from the base model and shows that KL timing matters: early regularization is more effective than late regularization, and sweeping the regularization strength reveals an empirical trade-off between task learning a...
A. Yanagisawa, B. Gho, Rajendran Ramesh Babu Manoj Narender et al.
· 0 citations