Abstract:The increasing integration of large language models (LLMs) into mental health applications necessitates robust frameworks for evaluating professional safety alignment. Current evaluative approaches primarily rely on refusal-based safety signals, which offer limited insight into the nuanced behaviors required in clinical practice. In mental health, clinically inadequate refusals can be perceived as unempathetic and discourage help-seeking. To address this gap, we move beyond refusal-centric metrics and introduce \texttt{PsychEthicsBench}, the first principle-grounded benchmark based on Australian psychology and psychiatry guidelines, designed to evaluate LLMs' ethical knowledge and behavioral responses through multiple-choice and open-ended tasks with fine-grained ethicality annotations. Empirical results across 14 models reveal that refusal rates are poor indicators of ethical behavior, revealing a significant divergence between safety triggers and clinical appropriateness. Notably, we find that domain-specific fine-tuning can degrade ethical robustness, as several specialized models underperform their base backbones in ethical alignment. PsychEthicsBench provides a foundation for systematic, jurisdiction-aware evaluation of LLMs in mental health, encouraging more responsible development in this domain.




Abstract:Seasonal time series Forecasting remains a challenging problem due to the long-term dependency from seasonality. In this paper, we propose a two-stage framework to forecast univariate seasonal time series. The first stage explicitly learns the long-range time series structure in a time window beyond the forecast horizon. By incorporating the learned long-range structure, the second stage can enhance the prediction accuracy in the forecast horizon. In both stages, we integrate the auto-regressive model with neural networks to capture both linear and non-linear characteristics in time series. Our framework achieves state-of-the-art performance on M4 Competition Hourly datasets. In particular, we show that incorporating the intermediate results generated in the first stage to existing forecast models can effectively enhance their prediction performance.