In recent years, large language models (LLMs) have made remarkable advancements, yet hallucination, where models produce inaccurate or non-factual statements, remains a significant challenge for real-world deployment. Although current classification-based methods, such as SAPLMA, are highly efficient in mitigating hallucinations, they struggle when non-factual information arises in the early or mid-sequence of outputs, reducing their reliability. To address these issues, we propose Hallucination Detection-Neural Differential Equations (HD-NDEs), a novel method that systematically assesses the truthfulness of statements by capturing the full dynamics of LLMs within their latent space. Our approaches apply neural differential equations (Neural DEs) to model the dynamic system in the latent space of LLMs. Then, the sequence in the latent space is mapped to the classification space for truth assessment. The extensive experiments across five datasets and six widely used LLMs demonstrate the effectiveness of HD-NDEs, especially, achieving over 14% improvement in AUC-ROC on the True-False dataset compared to state-of-the-art techniques.

The rapid advancement of vision-language models (VLMs) has brought a lot of attention to their safety alignment. However, existing methods have primarily focused on model undersafety, where the model responds to hazardous queries, while neglecting oversafety, where the model refuses to answer safe queries. In this paper, we introduce the concept of safety calibration, which systematically addresses both undersafety and oversafety. Specifically, we present VSCBench, a novel dataset of 3,600 image-text pairs that are visually or textually similar but differ in terms of safety, which is designed to evaluate safety calibration across image-centric and text-centric scenarios. Based on our benchmark, we evaluate safety calibration across eleven widely used VLMs. Our extensive experiments revealed major issues with both undersafety and oversafety. We further investigated four approaches to improve the model’s safety calibration. We found that even though some methods effectively calibrated the models’ safety problems, these methods also lead to the degradation of models’ utility. This trade-off underscores the urgent need for advanced calibration methods, and our benchmark provides a valuable tool for evaluating future approaches.

Large vision-language models (LVLMs) have made significant progress in recent years. While LVLMs exhibit excellent ability in language understanding, question answering, and conversations of visual inputs, they are prone to producing hallucinations. While several methods are proposed to evaluate the hallucinations in LVLMs, most are reference-based and depend on external tools, which complicates their practical application. To assess the viability of alternative methods, it is critical to understand whether the reference-free approaches, which do not rely on any external tools, can efficiently detect hallucinations. Therefore, we initiate an exploratory study to demonstrate the effectiveness of different reference-free solutions in detecting hallucinations in LVLMs. In particular, we conduct an extensive study on three kinds of techniques: uncertainty-based, consistency-based, and supervised uncertainty quantification methods on four representative LVLMs across two different tasks. The empirical results show that the reference-free approaches are capable of effectively detecting non-factual responses in LVLMs, with the supervised uncertainty quantification method outperforming the others, achieving the best performance across different settings.

Pre-trained language models are known to inherit a plethora of contextual biases from their training data. These biases have proven to be projected onto a variety of downstream applications, making their detection and mitigation imminent. Limited research has been conducted to quantify specific bias types, such as benevolent sexism, which may be subtly present within the inferred connotations of a sentence. To this extent, our work aims to: (1) provide a benchmark of sexism sentences; (2) adapt two bias metrics: mean probability score and mean normalized rank; (3) conduct a case study to quantify and analyze sexism in base and de-biased masked language models. We find that debiasing, even in its most effective form (Auto-Debias), solely nullifies the probability score of biasing tokens, while retaining them in high ranks. Auto-Debias illustrates a 90%-96% reduction in mean probability scores from base to debiased models, while only a 3%-16% reduction in mean normalized ranks. Similar to the application of non-parametric statistical tests for data that does not follow a normal distribution, operating on the ranks of predictions rather than their probability scores offers a more representative bias measure.