2 papers
cs.LG2026
SAGA: Score-Weighted Adaptive Generation Alignment for Low-Resource Nordic Language Models
Hoda Fakharzadehjahromy, Emil Wiman, Andreas Bueff +2
Preference optimisation has proven effective for improving large language models but typically relies on costly human preference annotations. Extending these methods to morphologic…
cs.LG2024
Fair4Free: Generating High-fidelity Fair Synthetic Samples using Data Free Distillation
Md Fahim Sikder, Daniel de Leng, Fredrik Heintz
This work presents Fair4Free, a novel generative model to generate synthetic fair data using data-free distillation in the latent space. Fair4Free can work on the situation when th…