Recent advancements in massively multilingual machine translation systems have significantly enhanced translation accuracy; however, even the best performing systems still generate hallucinations, severely impacting user trust. Detecting hallucinations in Machine Translation (MT) remains a critical challenge, particularly since existing methods excel with High-Resource Languages (HRLs) but exhibit substantial limitations when applied to Low-Resource Languages (LRLs). This paper evaluates hallucination detection approaches using Large Language Models (LLMs) and semantic similarity within massively multilingual embeddings. Our study spans 16 language directions, covering HRLs, LRLs, with diverse scripts. We find that the choice of model is essential for performance. On average, for HRLs, Llama3-70B outperforms the previous state of the art by as much as 0.16 MCC (Matthews Correlation Coefficient). However, for LRLs we observe that Claude Sonnet outperforms other LLMs on average by 0.03 MCC. The key takeaway from our study is that LLMs can achieve performance comparable or even better than previously proposed models, despite not being explicitly trained for any machine translation task. However, their advantage is less significant for LRLs.

Wee Siong Ng 1 wsng@i2r.a-star.edu.sg 1 Institute for Infocomm Research (I2R), A*ST AR, Singapore 2 School of Computing, National University of Singapore, Singapore Abstract Contrastive Language-Image Pre-training (CLIP) achieves remarkable performance in various downstream tasks through the alignment of image and text input embeddings and holds great promise for anomaly detection. However, our empirical experiments show that the embeddings of text inputs unexpectedly tightly cluster together, far away from image embeddings, contrary to the model's contrastive training objective to align image-text input pairs. We show that this phenomenon induces a'similarity bias' - in which false negative and false positive errors occur due to bias in the similarities between images and the normal label text embeddings. To address this bias, we propose a novel methodology called BLISS which directly accounts for this similarity bias through the use of an auxiliary, external set of text inputs. BLISS is simple, it does not require strong inductive biases about anomalous behaviour nor an expensive training process, and it significantly outperforms baseline methods on benchmark image datasets, even when access to normal data is extremely limited. 1 Introduction Anomaly detection (AD) is an important task in many vision-related applications, such as medical diagnosis and industrial defect detection. Neural networks are trained to embed input images into a latent space where anomalies are more easily detected. Vision-language models (VLM), which embed both image and textual inputs, have surged in popularity recently due to their flexibility and strong performance in various downstream tasks.

Antimicrobial Resistance represents a significant challenge in the Intensive Care Unit (ICU), where patients are at heightened risk of Multidrug-Resistant (MDR) infections-pathogens resistant to multiple antimicrobial agents. This study introduces a novel methodology that integrates Gated Recurrent Units (GRUs) with advanced intrinsic and post-hoc interpretability techniques for detecting the onset of MDR in patients across time. Within interpretability methods, we propose Explainable Artificial Intelligence (XAI) approaches to handle irregular Multivariate Time Series (MTS), introducing Irregular Time Shapley Additive Explanations (IT-SHAP), a modification of Shapley Additive Explanations designed for irregular MTS with Recurrent Neural Networks focused on temporal outputs. Our methodology aims to identify specific risk factors associated with MDR in ICU patients. GRU with Hadamard's attention demonstrated high initial specificity and increasing sensitivity over time, correlating with increased nosocomial infection risks during prolonged ICU stays. XAI analysis, enhanced by Hadamard attention and IT-SHAP, identified critical factors such as previous non-resistant cultures, specific antibiotic usage patterns, and hospital environment dynamics. These insights suggest that early detection of at-risk patients can inform interventions such as preventive isolation and customized treatments, significantly improving clinical outcomes. The proposed GRU model for temporal classification achieved an average Receiver Operating Characteristic Area Under the Curve of 78.27 +- 1.26 over time, indicating strong predictive performance. In summary, this study highlights the clinical utility of our methodology, which combines predictive accuracy with interpretability, thereby facilitating more effective healthcare interventions by professionals.

We propose a new method for feature learning and function estimation in supervised learning via regularised empirical risk minimisation. Our approach considers functions as expectations of Sobolev functions over all possible one-dimensional projections of the data. This framework is similar to kernel ridge regression, where the kernel is $\mathbb{E}_w ( k^{(B)}(w^\top x,w^\top x^\prime))$, with $k^{(B)}(a,b) := \min(|a|, |b|)1_{ab>0}$ the Brownian kernel, and the distribution of the projections $w$ is learnt. This can also be viewed as an infinite-width one-hidden layer neural network, optimising the first layer's weights through gradient descent and explicitly adjusting the non-linearity and weights of the second layer. We introduce an efficient computation method for the estimator, called Brownian Kernel Neural Network (BKerNN), using particles to approximate the expectation. The optimisation is principled due to the positive homogeneity of the Brownian kernel. Using Rademacher complexity, we show that BKerNN's expected risk converges to the minimal risk with explicit high-probability rates of $O( \min((d/n)^{1/2}, n^{-1/6}))$ (up to logarithmic factors). Numerical experiments confirm our optimisation intuitions, and BKerNN outperforms kernel ridge regression, and favourably compares to a one-hidden layer neural network with ReLU activations in various settings and real data sets.

General-purpose artificial intelligence (AI) systems are built on massive swathes of public web data, assembled into corpora such as C4, RefinedWeb, and Dolma. To our knowledge, we conduct the first, large-scale, longitudinal audit of the consent protocols for the web domains underlying AI training corpora. Our audit of 14,000 web domains provides an expansive view of crawlable web data and how codified data use preferences are changing over time. We observe a proliferation of AI-specific clauses to limit use, acute differences in restrictions on AI developers, as well as general inconsistencies between websites' expressed intentions in their Terms of Service and their robots.txt. We diagnose these as symptoms of ineffective web protocols, not designed to cope with the widespread re-purposing of the internet for AI. Our longitudinal analyses show that in a single year (2023-2024) there has been a rapid crescendo of data restrictions from web sources, rendering ~5%+ of all tokens in C4, or 28%+ of the most actively maintained, critical sources in C4, fully restricted from use. For Terms of Service crawling restrictions, a full 45% of C4 is now restricted. If respected or enforced, these restrictions are rapidly biasing the diversity, freshness, and scaling laws for general-purpose AI systems. We hope to illustrate the emerging crises in data consent, for both developers and creators. The foreclosure of much of the open web will impact not only commercial AI, but also non-commercial AI and academic research.

Recent advancements in natural language processing (NLP) have led to automation in various domains. However, clinical NLP often relies on benchmark datasets that may not reflect real-world scenarios accurately. Automatic ICD coding, a vital NLP task, typically uses outdated and imbalanced datasets like MIMIC-III, with existing methods yielding micro-averaged F1 scores between 0.4 and 0.7 due to many false positives. Our research introduces an enhanced approach to ICD coding that improves F1 scores by using chapter-based named entities and attentional models. This method categorizes discharge summaries into ICD-9 Chapters and develops attentional models with chapter-specific data, eliminating the need to consider external data for code identification. For categorization, we use Chapter-IV to de-bias and influence key entities and weights without neural networks, creating accurate thresholds and providing interpretability for human validation. Post-validation, we develop attentional models for three frequent and three non-frequent codes from Chapter-IV using Bidirectional-Gated Recurrent Units (GRUs) with Attention and Transformer with Multi-head Attention architectures. The average Micro-F1 scores of 0.79 and 0.81 from these models demonstrate significant performance improvements in ICD coding.

This study introduces novel superior scoring rules called Penalized Brier Score (PBS) and Penalized Logarithmic Loss (PLL) to improve model evaluation for probabilistic classification. Traditional scoring rules like Brier Score and Logarithmic Loss sometimes assign better scores to misclassifications in comparison with correct classifications. This discrepancy from the actual preference for rewarding correct classifications can lead to suboptimal model selection. By integrating penalties for misclassifications, PBS and PLL modify traditional proper scoring rules to consistently assign better scores to correct predictions. Formal proofs demonstrate that PBS and PLL satisfy strictly proper scoring rule properties while also preferentially rewarding accurate classifications. Experiments showcase the benefits of using PBS and PLL for model selection, model checkpointing, and early stopping. PBS exhibits a higher negative correlation with the F1 score compared to the Brier Score during training. Thus, PBS more effectively identifies optimal checkpoints and early stopping points, leading to improved F1 scores. Comparative analysis verifies models selected by PBS and PLL achieve superior F1 scores. Therefore, PBS and PLL address the gap between uncertainty quantification and accuracy maximization by encapsulating both proper scoring principles and explicit preference for true classifications. The proposed metrics can enhance model evaluation and selection for reliable probabilistic classification.

Predictive machine learning models are becoming increasingly deployed in high-stakes contexts involving sensitive personal data; in these contexts, there is a trade-off between model explainability and data privacy. In this work, we push the boundaries of this trade-off: with a focus on foundation models for image classification fine-tuning, we reveal unforeseen privacy risks of post-hoc model explanations and subsequently offer mitigation strategies for such risks. First, we construct VAR-LRT and L1/L2-LRT, two new membership inference attacks based on feature attribution explanations that are significantly more successful than existing explanation-leveraging attacks, particularly in the low false-positive rate regime that allows an adversary to identify specific training set members with confidence. Second, we find empirically that optimized differentially private fine-tuning substantially diminishes the success of the aforementioned attacks, while maintaining high model accuracy. We carry out a systematic empirical investigation of our 2 new attacks with 5 vision transformer architectures, 5 benchmark datasets, 4 state-of-the-art post-hoc explanation methods, and 4 privacy strength settings.

Initially, IT operations were predominantly manual, relying heavily on human intervention for system monitoring, troubleshooting, and problem resolution. However, with the escalating scale and complexity of systems, the efficacy and precision of manual operations have been increasingly challenged. Subsequently, DevOps was introduced, building upon manual operations and fostering a synergistic collaboration between development and operations. Through automated deployment and continuous integration, DevOps has the capability to expedite the release of new features and rectify issues with greater speed and reliability. Nonetheless, DevOps still necessitates manual involvement in certain complex decision-making processes and tasks. To further mitigate this challenge and enhance cost-effectiveness and efficiency, AIOps leverages machine learning and data analysis to automatically collect and scrutinize vast amounts of IT operation data, enabling real-time monitoring, anomaly detection, fault localization, and automated processing of IT systems. AIOps not only augments the efficiency and accuracy of IT operations but also equips IT operations with the capacity to adapt more effectively to complex and dynamic IT environments, utilizing artificial intelligence and big data technologies.

Despite the evolution of language models, they continue to portray harmful societal biases and stereotypes inadvertently learned from training data. These inherent biases often result in detrimental effects in various applications. Counterfactual Data Augmentation (CDA), which seeks to balance demographic attributes in training data, has been a widely adopted approach to mitigate bias in natural language processing. However, many existing CDA approaches rely on word substitution techniques using manually compiled word-pair dictionaries. These techniques often lead to out-of-context substitutions, resulting in potential quality issues. The advancement of model-based techniques, on the other hand, has been challenged by the need for parallel training data. Works in this area resort to manually generated parallel data that are expensive to collect and are consequently limited in scale. This paper proposes FairFlow, an automated approach to generating parallel data for training counterfactual text generator models that limits the need for human intervention. Furthermore, we show that FairFlow significantly overcomes the limitations of dictionary-based word-substitution approaches whilst maintaining good performance.