Search papers, labs, and topics across Lattice.
National Institute of Informatics
2
0
5
DeSRPA achieves superior emotional consistency and speech naturalness in SRPAs without the need for extensive retraining, challenging the dominance of end-to-end fine-tuning approaches.
Achieve stable and interpretable persona control in LLMs by routing contrastively-trained personality vectors aligned with the Big Five facets, outperforming prompt engineering and other baselines.
