Helpfulness Hurts: Domain-Dependent Degradation of Mid-Trained Compassion Values Under Post-Training 文章

ArXiv CS.CL2026-07-14PAPERen作者: Jasmine Brazilek, Juliana Seawell

Helpfulness Hurts: Domain-Dependent Degradation of Mid-Trained Compassion Values Under Post-Training · 相关技术