Data-driven decision support tools play an increasingly central role in decision-making across various domains. In this work, we focus on binary classification models for predicting positive-outcome scores and deciding on resource allocation, e.g., credit scores for granting loans or churn propensity scores for targeting customers with a retention campaign. Such models may exhibit discriminatory behavior toward specific demographic groups through their predicted scores, potentially leading to unfair resource allocation. We focus on demographic parity as a fairness metric to compare the proportions of instances that are selected based on their positive outcome scores across groups. In this work, we propose a decision-centric fairness methodology that induces fairness only within the decision-making region -- the range of relevant decision thresholds on the score that may be used to decide on resource allocation -- as an alternative to a global fairness approach that seeks to enforce parity across the entire score distribution. By restricting the induction of fairness to the decision-making region, the proposed decision-centric approach avoids imposing overly restrictive constraints on the model, which may unnecessarily degrade the quality of the predicted scores. We empirically compare our approach to a global fairness approach on multiple (semi-synthetic) datasets to identify scenarios in which focusing on fairness where it truly matters, i.e., decision-centric fairness, proves beneficial.

We present ReasonIR-8B, the first retriever specifically trained for general reasoning tasks. Existing retrievers have shown limited gains on reasoning tasks, in part because existing training datasets focus on short factual queries tied to documents that straightforwardly answer them. We develop a synthetic data generation pipeline that, for each document, our pipeline creates a challenging and relevant query, along with a plausibly related but ultimately unhelpful hard negative. By training on a mixture of our synthetic data and existing public data, ReasonIR-8B achieves a new state-of-the-art of 29.9 nDCG@10 without reranker and 36.9 nDCG@10 with reranker on BRIGHT, a widely-used reasoning-intensive information retrieval (IR) benchmark. When applied to RAG tasks, ReasonIR-8B improves MMLU and GPQA performance by 6.4% and 22.6% respectively, relative to the closed-book baseline, outperforming other retrievers and search engines. In addition, ReasonIR-8B uses test-time compute more effectively: on BRIGHT, its performance consistently increases with longer and more information-rich rewritten queries; it continues to outperform other retrievers when combined with an LLM reranker. Our training recipe is general and can be easily extended to future LLMs; to this end, we open-source our code, data, and model.

Previous detoxification methods are typically model-specific, addressing only individual models or model families, and require careful hyperparameter tuning due to the trade-off between detoxification efficacy and language modeling performance. Specifically, we propose a novel and efficient dataset distillation technique for detoxification using contrastive decoding. This approach distills detoxifying representations in the form of synthetic text data, enabling universal detoxification of any LLM through fine-tuning with the distilled text. Our experiments demonstrate that the detoxifying text distilled from GPT -2 can effectively detoxify larger models, including OPT, Falcon, and LLaMA-2. Additionally, analysis of the detoxifying text reveals a reduction in politically biased content, providing insights into the attributes necessary for effective detoxification of LLMs. Our codes are available at https://github.com/EminLU/UniDetox. Fascinated by the remarkable capabilities of Large Language Models (LLMs), numerous researchers and developers are dedicating their efforts to building new models. Today, many off-the-shelf pre-trained LLMs are publicly available (Radford et al., 2019; Zhang et al., 2022; Almazrouei et al., 2023; Touvron et al., 2023), and practitioners employ them in a wide range of applications. While this trend is expected to drive innovation across various fields, it simultaneously raises significant concerns regarding the unintended harmful behaviors exhibited by LLMs. LLMs, developed through pre-training on a large-scale corpus, often unintentionally acquire toxic content present in their training datasets (Gehman et al., 2020; Webster et al., 2020; Nozza et al., 2021). Due to these concerns, there have been efforts to introduce comprehensive regulations to mitigate the toxicity of LLMs; however, there is currently no standardized approach capable of consistently removing toxic content across diverse models. By developing a universal detoxification approach, we can form the basis for broadly applicable regulations and ensure consistent toxicity mitigation across a wide variety of LLMs. While numerous studies have explored the detoxification of LLMs, there is currently no post-hoc approach that can be seamlessly applied across models with varying architectures, sizes, or tok-enizers. Existing post-hoc detoxification strategies include decoding-time control (Liu et al., 2021; Zhang & Wan, 2023), word embedding/logits modification (Gehman et al., 2020; Han et al., 2024), and model editing (Ilharco et al., 2023; Wang et al., 2024). Crucially, this equilibrium point varies across models, necessitating individual hyperparameter optimization for each model, as we will thoroughly investigate in our experiments.

In arbitrary-order language models, it is an open question how to sample tokens in parallel from the correct joint distribution. With discrete diffusion models, the more tokens they generate in parallel, the less their predicted distributions adhere to the originally learned data distribution, as they rely on a conditional independence assumption that only works with infinitesimally small timesteps. We find that a different class of models, any-subset autoregressive models (AS-ARMs), holds the solution. As implied by the name, AS-ARMs can generate tokens in any order, and in parallel. Moreover, AS-ARMs support parallelized joint probability density estimation, allowing them to correct their own parallel-generated token distributions, via our Any-Subset Speculative Decoding (ASSD) algorithm. ASSD provably enables generation of tokens from the correct joint distribution, with the number of neural network calls upper bounded by the number of tokens predicted. We empirically verify that ASSD speeds up language generation, without sacrificing quality. Furthermore, we provide a mathematically justified scheme for training AS-ARMs for generation, and show that AS-ARMs achieve state-of-the-art performance among sub-200M parameter models on infilling benchmark tasks, and nearly match the performance of models 50X larger on code generation. Our theoretical and empirical results indicate that the once-forgotten AS-ARMs are a promising direction of language modeling.

In this work, we theoretically demonstrate that current graph positional encodings (PEs) are not beneficial and could potentially hurt performance in tasks involving heterophilous graphs, where nodes that are close tend to have different labels. This limitation is critical as many real-world networks exhibit heterophily, and even highly homophilous graphs can contain local regions of strong heterophily. To address this limitation, we propose Learnable Laplacian Positional Encodings (LLPE), a new PE that leverages the full spectrum of the graph Laplacian, enabling them to capture graph structure on both homophilous and heterophilous graphs. Theoretically, we prove LLPE's ability to approximate a general class of graph distances and demonstrate its generalization properties. Empirically, our evaluation on 12 benchmarks demonstrates that LLPE improves accuracy across a variety of GNNs, including graph transformers, by up to 35% and 14% on synthetic and real-world graphs, respectively. Going forward, our work represents a significant step towards developing PEs that effectively capture complex structures in heterophilous graphs.

Homophily, as a measure, has been critical to increasing our understanding of graph neural networks (GNNs). However, to date this measure has only been analyzed in the context of static graphs. In our work, we explore homophily in dynamic settings. Focusing on graph convolutional networks (GCNs), we demonstrate theoretically that in dynamic settings, current GCN discriminative performance is characterized by the probability that a node's future label is the same as its neighbors' current labels. Based on this insight, we propose dynamic homophily, a new measure of homophily that applies in the dynamic setting. This new measure correlates with GNN discriminative performance and sheds light on how to potentially design more powerful GNNs for dynamic graphs. Leveraging a variety of dynamic node classification datasets, we demonstrate that popular GNNs are not robust to low dynamic homophily. Going forward, our work represents an important step towards understanding homophily and GNN performance in dynamic node classification.

Digital twins (DTs) are improving water distribution systems by using real-time data, analytics, and prediction models to optimize operations. This paper presents a DT platform designed for a Spanish water supply network, utilizing Long Short-Term Memory (LSTM) networks to predict water consumption. However, machine learning models are vulnerable to adversarial attacks, such as the Fast Gradient Sign Method (FGSM) and Projected Gradient Descent (PGD). These attacks manipulate critical model parameters, injecting subtle distortions that degrade forecasting accuracy. To further exploit these vulnerabilities, we introduce a Learning Automata (LA) and Random LA-based approach that dynamically adjusts perturbations, making adversarial attacks more difficult to detect. Experimental results show that this approach significantly impacts prediction reliability, causing the Mean Absolute Percentage Error (MAPE) to rise from 26% to over 35%. Moreover, adaptive attack strategies amplify this effect, highlighting cybersecurity risks in AI-driven DTs. These findings emphasize the urgent need for robust defenses, including adversarial training, anomaly detection, and secure data pipelines.

--Water distribution systems in rural areas face serious challenges such as a lack of real-time monitoring, vulnerability to cyberattacks, and unreliable data handling. This paper presents an integrated framework that combines LoRaW ANbased data acquisition, a machine learning-driven Intrusion Detection System (IDS), and a blockchain-enabled Digital Twin (BC-DT) platform for secure and transparent water management. The IDS filters anomalous or spoofed data using a Long Short-T erm Memory (LSTM) Autoencoder and Isolation Forest before validated data is logged via smart contracts on a private Ethereum blockchain using Proof of Authority (PoA) consensus. The verified data feeds into a real-time DT model supporting leak detection, consumption forecasting, and predictive maintenance. Experimental results demonstrate that the system achieves over 80 transactions per second (TPS) with under 2 seconds of latency while remaining cost-effective and scalable for up to 1,000 smart meters. This work demonstrates a practical and secure architecture for decentralized water infrastructure in under-connected rural environments. While remote-sensing techniques have proven valuable for water quality monitoring [1], [2], regarding water distribution, efficient distribution is a significant issue in rural regions, especially where infrastructure is poor and digital monitoring is scarce.

--Digital twins (DTs) help improve real-time monitoring and decision-making in water distribution systems. However, their connectivity makes them easy targets for cyberattacks such as scanning, denial-of-service (DoS), and unauthorized access. Small and medium-sized enterprises (SMEs) that manage these systems often do not have enough budget or staff to build strong cybersecurity teams. T o solve this problem, we present a Virtual Cybersecurity Department (VCD), an affordable and automated framework designed for SMEs. The VCD uses open-source tools like Zabbix for real-time monitoring, Suricata for network intrusion detection, Fail2Ban to block repeated login attempts, and simple firewall settings. T o improve threat detection, we also add a machine-learning-based IDS trained on the OD-IDS2022 dataset using an improved ensemble model. Our solution gives SMEs a practical and efficient way to secure water systems using low-cost and easy-to-manage tools.

A key factor for lunar mission planning is the ability to assess the local availability of raw materials. However, many potentially relevant measurements are scattered across a variety of scientific publications. In this paper we consider the viability of obtaining lunar composition data by leveraging LLMs to rapidly process a corpus of scientific publications. While leveraging LLMs to obtain knowledge from scientific documents is not new, this particular application presents interesting challenges due to the heterogeneity of lunar samples and the nuances involved in their characterization. Accuracy and uncertainty quantification are particularly crucial since many materials properties can be sensitive to small variations in composition. Our findings indicate that off-the-shelf LLMs are generally effective at extracting data from tables commonly found in these documents. However, there remains opportunity to further refine the data we extract in this initial approach; in particular, to capture fine-grained mineralogy information and to improve performance on more subtle/complex pieces of information.