Algorithms are increasingly used to automate or aid human decisions, yet recent research shows that these algorithms may exhibit bias across legally protected demographic groups. However, data on these groups may be unavailable to organizations or external auditors due to privacy legislation. This paper studies bias detection using an unsupervised clustering tool when data on demographic groups are unavailable. We collaborate with the Dutch Executive Agency for Education to audit an algorithm that was used to assign risk scores to college students at the national level in the Netherlands between 2012-2023. Our audit covers more than 250,000 students from the whole country. The unsupervised clustering tool highlights known disparities between students with a non-European migration background and Dutch origin. Our contributions are three-fold: (1) we assess bias in a real-world, large-scale and high-stakes decision-making process by a governmental organization; (2) we use simulation studies to highlight potential pitfalls of using the unsupervised clustering tool to detect true bias when demographic group data are unavailable and provide recommendations for valid inferences; (3) we provide the unsupervised clustering tool in an open-source library. Our work serves as a starting point for a deliberative assessment by human experts to evaluate potential discrimination in algorithmic-supported decision-making processes.

Large language models (LLMs) have emerged as powerful tools for addressing a wide range of general inquiries and tasks. Despite this, fine-tuning aligned LLMs on smaller, domain-specific datasets, critical to adapting them to specialized tasks, can inadvertently degrade their safety alignment, even when the datasets are benign. This phenomenon makes models more susceptible to providing inappropriate responses. In this study, we systematically examine the factors contributing to safety alignment degradation in benign fine-tuning scenarios. Our analysis identifies three critical factors affecting aligned LLMs: answer structure, identity calibration, and role-play. Additionally, we evaluate the reliability of state-of-the-art reward models (RMs), which are often used to guide alignment processes. Our findings reveal that these RMs frequently fail to accurately reflect human preferences regarding safety, underscoring their limitations in practical applications. By uncovering these challenges, our work highlights the complexities of maintaining safety alignment during fine-tuning and offers guidance to help developers balance utility and safety in LLMs. Datasets and fine-tuning code used in our experiments can be found in https://github.com/GuanlinLee/llm_instruction_tuning.

Anomaly detection is essential for identifying rare and significant events across diverse domains such as finance, cybersecurity, and network monitoring. This paper presents Synthetic Anomaly Monitoring (SAM), an innovative approach that applies synthetic control methods from causal inference to improve both the accuracy and interpretability of anomaly detection processes. By modeling normal behavior through the treatment of each feature as a control unit, SAM identifies anomalies as deviations within this causal framework. We conducted extensive experiments comparing SAM with established benchmark models, including Isolation Forest, Local Outlier Factor (LOF), k-Nearest Neighbors (kNN), and One-Class Support Vector Machine (SVM), across five diverse datasets, including Credit Card Fraud, HTTP Dataset CSIC 2010, and KDD Cup 1999, among others. Our results demonstrate that SAM consistently delivers robust performance, highlighting its potential as a powerful tool for real-time anomaly detection in dynamic and complex environments.

Detecting hallucinations in Large Language Models (LLMs) remains a critical challenge for their reliable deployment in real-world applications. To address this, we introduce SelfCheckAgent, a novel framework integrating three different agents: the Symbolic Agent, the Specialized Detection Agent, and the Contextual Consistency Agent. These agents provide a robust multi-dimensional approach to hallucination detection. Notable results include the Contextual Consistency Agent leveraging Llama 3.1 with Chain-of-Thought (CoT) to achieve outstanding performance on the WikiBio dataset, with NonFactual hallucination detection scoring 93.64%, Factual 70.26%, and Ranking 78.48% respectively. On the AIME dataset, GPT-4o with CoT excels in NonFactual detection with 94.89% but reveals trade-offs in Factual with 30.58% and Ranking with 30.68%, underscoring the complexity of hallucination detection in the complex mathematical domains. The framework also incorporates a triangulation strategy, which increases the strengths of the SelfCheckAgent, yielding significant improvements in real-world hallucination identification. The comparative analysis demonstrates SelfCheckAgent's applicability across diverse domains, positioning it as a crucial advancement for trustworthy LLMs. These findings highlight the potentiality of consistency-driven methodologies in detecting hallucinations in LLMs.

We consider a general model for high-dimensional empirical risk minimization whereby the data $\mathbf{x}_i$ are $d$-dimensional isotropic Gaussian vectors, the model is parametrized by $\mathbf{\Theta}\in\mathbb{R}^{d\times k}$, and the loss depends on the data via the projection $\mathbf{\Theta}^\mathsf{T}\mathbf{x}_i$. This setting covers as special cases classical statistics methods (e.g. multinomial regression and other generalized linear models), but also two-layer fully connected neural networks with $k$ hidden neurons. We use the Kac-Rice formula from Gaussian process theory to derive a bound on the expected number of local minima of this empirical risk, under the proportional asymptotics in which $n,d\to\infty$, with $n\asymp d$. Via Markov's inequality, this bound allows to determine the positions of these minimizers (with exponential deviation bounds) and hence derive sharp asymptotics on the estimation and prediction error. In this paper, we apply our characterization to convex losses, where high-dimensional asymptotics were not (in general) rigorously established for $k\ge 2$. We show that our approach is tight and allows to prove previously conjectured results. In addition, we characterize the spectrum of the Hessian at the minimizer. A companion paper applies our general result to non-convex examples.

Contrastive Language-Image Pre-training (CLIP) has become the standard for cross-modal image-text representation learning. Improving CLIP typically requires additional data and retraining with new loss functions, but these demands raise resource and time costs, limiting practical use. In this work, we introduce HELIP, a cost-effective strategy that improves CLIP models by exploiting challenging text-image pairs within existing datasets in continuous training. This eliminates the need for additional data or extensive retraining. Moreover, HELIP integrates effortlessly into current training pipelines with minimal code modifications, allowing for quick and seamless implementation. On comprehensive benchmarks, HELIP consistently boosts existing models. In particular, within just two epochs of training, it improves zero-shot classification accuracy on ImageNet for SLIP models pre-trained on CC3M, CC12M, and YFCC15M datasets by 3.05%, 4.47%, and 10.1% , respectively. In addition, on fine-grained classification datasets, HELIP improves the zero-shot performance of CLIP and SLIP by an average of 8.4% and 18.6%, and their linear probe performance by an average of 9.5% and 3.0%. The code is publicly available at: https://github.com/haonan3/HELIP-NACCL-2025.git.

Since the explosion in popularity of ChatGPT, large language models (LLMs) have continued to impact our everyday lives. Equipped with external tools that are designed for a specific purpose (e.g., for flight booking or an alarm clock), LLM agents exercise an increasing capability to assist humans in their daily work. Although LLM agents have shown a promising blueprint as daily assistants, there is a limited understanding of how they can provide daily assistance based on planning and sequential decision making capabilities. We draw inspiration from recent work that has highlighted the value of 'LLM-modulo' setups in conjunction with humans-in-the-loop for planning tasks. We conducted an empirical study (N = 248) of LLM agents as daily assistants in six commonly occurring tasks with different levels of risk typically associated with them (e.g., flight ticket booking and credit card payments). To ensure user agency and control over the LLM agent, we adopted LLM agents in a plan-then-execute manner, wherein the agents conducted step-wise planning and step-by-step execution in a simulation environment. We analyzed how user involvement at each stage affects their trust and collaborative team performance. Our findings demonstrate that LLM agents can be a double-edged sword -- (1) they can work well when a high-quality plan and necessary user involvement in execution are available, and (2) users can easily mistrust the LLM agents with plans that seem plausible. We synthesized key insights for using LLM agents as daily assistants to calibrate user trust and achieve better overall task outcomes. Our work has important implications for the future design of daily assistants and human-AI collaboration with LLM agents.

Achieving consensus in group decision-making often involves overcoming significant challenges, particularly in reconciling diverse perspectives and mitigating biases that hinder agreement. Traditional methods relying on human facilitators are often constrained by scalability and efficiency, especially in large-scale, fast-paced discussions. To address these challenges, this study proposes a novel framework employing large language models (LLMs) as automated facilitators within a custom-built multi-user chat system. Leveraging cosine similarity as a core metric, this approach evaluates the ability of three state-of-the-art LLMs- ChatGPT 4.0, Mistral Large 2, and AI21 Jamba Instruct- to synthesize consensus proposals that align with participants' viewpoints. Unlike conventional techniques, the system integrates adaptive facilitation strategies, including clarifying misunderstandings, summarizing discussions, and proposing compromises, enabling the LLMs to iteratively refine consensus proposals based on user feedback. Experimental results demonstrate the superiority of ChatGPT 4.0, which achieves higher alignment with participant opinions, requiring fewer iterations to reach consensus compared to its counterparts. Moreover, analysis reveals the nuanced performance of the models across various sustainability-focused discussion topics, such as climate action, quality education, good health and well-being, and access to clean water and sanitation. These findings highlight the transformative potential of LLM-driven facilitation for improving collective decision-making processes and underscore the importance of advancing evaluation metrics and cross-cultural adaptability in future research.

Technical standards, or simply standards, are established documented guidelines and rules that facilitate the interoperability, quality, and accuracy of systems and processes. In recent years, we have witnessed an emerging paradigm shift where the adoption of generative AI (GenAI) models has increased tremendously, spreading implementation interests across standard-driven industries, including engineering, legal, healthcare, and education. In this paper, we assess the criticality levels of different standards across domains and sectors and complement them by grading the current compliance capabilities of state-of-the-art GenAI models. To support the discussion, we outline possible challenges and opportunities with integrating GenAI for standard compliance tasks while also providing actionable recommendations for entities involved with developing and using standards. Overall, we argue that aligning GenAI with standards through computational methods can help strengthen regulatory and operational compliance. We anticipate this area of research will play a central role in the management, oversight, and trustworthiness of larger, more powerful GenAI-based systems in the near future.

Recent technological advances have introduced new high-throughput methods for studying host-virus interactions, but testing synergistic interactions between host gene pairs during infection remains relatively slow and labor intensive. Identification of multiple gene knockdowns that effectively inhibit viral replication requires a search over the combinatorial space of all possible target gene pairs and is infeasible via brute-force experiments. Although active learning methods for sequential experimental design have shown promise, existing approaches have generally been restricted to single-gene knockdowns or small-scale double knockdown datasets. In this study, we present an integrated Deep Active Learning (DeepAL) framework that incorporates information from a biological knowledge graph (SPOKE, the Scalable Precision Medicine Open Knowledge Engine) to efficiently search the configuration space of a large dataset of all pairwise knockdowns of 356 human genes in HIV infection. Through graph representation learning, the framework is able to generate task-specific representations of genes while also balancing the exploration-exploitation trade-off to pinpoint highly effective double-knockdown pairs. We additionally present an ensemble method for uncertainty quantification and an interpretation of the gene pairs selected by our algorithm via pathway analysis. To our knowledge, this is the first work to show promising results on double-gene knockdown experimental data of appreciable scale (356 by 356 matrix).