Large language models (LLMs) are increasingly used to answer questions about political information, including in election-adjacent information settings where factual errors and ideological distortions are high-stakes. We present a reproducible measurement framework that treats hallucinations, unsupported statements in document-grounded QA, as diagnostic signals of ideological drift. Using 21,727 expert-labeled U.S. political news articles from QBias spanning left, center, and right sources, we (i) generate an article-specific question, (ii) elicit document-grounded answers from three open-weight LLMs and one proprietary model, (iii) detect sentence-level hallucinations via reference-based comparison, (iv) classify the ideological valence of hallucinated sentences with a fine-tuned stance classifier, and (v) probe output logits to relate token-level uncertainty to hallucination and drift. Hallucination rates vary substantially across models and concentrate in contentious topics, while source-ideology differences in hallucination frequency are modest. In contrast, hallucination content exhibits robust leftward drift: a majority of hallucinated sentences are classified as left-leaning, including among hallucinations generated from right-leaning sources. Logit-level analysis shows hallucinations arise in high-entropy generation contexts, and in some models uncertainty also predicts leftward drift, consistent with an "uncertainty → guessing" mechanism. We discuss implications for auditing AI-mediated political information and for designing safeguards in election-relevant deployments.
论文检索
输入标题、作者或关键词，从 100,903 篇学术成果中精准定位
Agent-based simulations have an untapped potential to inform social policies on urgent human development challenges in a non-invasive way, before these are implemented in real-world populations. This paper responds to the request from non-profit and governmental organizations to evaluate policies under discussion to improve equity in health care services for people experiencing homelessness (PEH) in the city of Barcelona. With this goal, we integrate the conceptual framework of the capability approach (CA), which is explicitly designed to promote and assess human well-being, to model and evaluate the behavior of agents who represent PEH and social workers. We define a reinforcement learning environment where agents aim to restore their central human capabilities, under existing environmental and legal constraints. We use Bayesian calibration of profile-dependent behavioral parameters of PEH agents, modeling the degree of trust and resulting engagement with social workers, which is reportedly a key element for the success of the policies in scope. Our results open a path to mitigate health inequity by building relationships of trust between social service workers and PEH.
Every day, many people die under violent circumstances, whether from crimes, war, migration, or climate disasters. Medico-legal and law enforcement institutions document many portraits of the deceased for evidence, but cannot immediately carry out identification on them. While traditional image editing tools can process these photos for public release, the workflow is lengthy and produces suboptimal results. In this work, we leverage advances in image generation models, which can now produce photorealistic human portraits, to introduce FlowID, an identity-preserving facial reconstruction method. Our approach combines single-image fine-tuning, which adapts the generative model to out-of-distribution injured faces, with attention-based masking that localizes edits to damaged regions while preserving identity-critical features. Together, these components enable the removal of artifacts from violent death while retaining sufficient identity information to support identification. To evaluate our method, we introduce InjuredFaces, a novel benchmark for identity-preserving facial reconstruction under severe facial damage. Beyond serving as an evaluation tool for this work, InjuredFaces provides a standardized resource for the community to study and compare methods addressing facial reconstruction in extreme conditions. Experimental results show that FlowID outperforms state-of-the-art open-source methods while maintaining low memory requirements, making it suitable for local deployment without compromising data privacy.
Diabetic retinopathy (DR) and diabetic macular edema (DME) are leading causes of preventable blindness worldwide. Automated screening tools are critical for timely detection at scale, particularly in low-resource settings where access to ophthalmologists is limited. We propose DRDME-Net, a deployment-driven joint learning framework that formulates DR grading as an ordinal regression task and DME detection via a continuous surrogate, rather than conventional classification. This design yields stable risk scores tightly aligned with operational clinical decision-making thresholds. Evaluation on facility and community cohorts demonstrates that DRDME-Net achieves strong performance across severity boundaries. Insights from an initial feasibility pilot further demonstrate its scalability in real-world workflows. These results highlight the potential of DRDME-Net to expand equitable access to timely detection, reduce preventable vision loss, and provide a practical template for integrating AI into population screening initiatives.
The proliferation of sophisticated telephone scams poses a significant societal and economic threat, impacting diverse linguistic contexts in a country like India. Furthermore, the lack of large-scale, publicly available datasets remains a critical barrier impacting research on robust, real-time countermeasures. In view of this, the proposed work introduces ICFD-31k, the first Indian Conversational Fraud Dataset, representing a new benchmark containing over 31,000 realistic conversational transcripts. ICFD-31k comprises systematically generated content, covering 10 distinct fraud umbrellas spanning from financial impersonation to job scams. ICFD-31k transcripts feature rich annotations comprising a final verdict, chunk-level streaming labels, and detailed ``slow-thinking'' rationales. In addition, the human-in-the-loop evaluation validates the ICFD-31k's quality, achieving a Cohen's Kappa of 0.534 that confirms annotation reliability. Furthermore, the proposed work introduces two fine-tuned models based on RoBERTa: M1 for non-streaming data and M2 for streaming data. The comprehensive experiments with strong baselines (M1, M2) further demonstrate the ICFD-31k's utility. The code and reproducibility materials are available at https://github.com/SPELLAILab/ICFD-31k.
In nuclear power plant accidents, airborne radioiodine poses major health risks, making reliable reconstructions of its spatiotemporal distribution crucial for emergency management. Current state-of-the-art prognosis systems use atmospheric dispersion modeling but ignore posterior evidence from emergency care centers, comprising movement profiles and thyroid measurements of affected individuals. A first study showed that the AI method Cuboid Splatting can reconstruct iodine air concentrations from such data but it ignores simulations from established prognosis systems. Our multidisciplinary team extends Cuboid Splatting by incorporating these simulations as priors and subsequently correcting them using movement and thyroid data. Several ways to translate and correct priors are developed. The best-performing approaches are combined into a novel Cuboid Splatting-with-prior mechanism, which we evaluate using constructed prior scenarios representing different error types and intensities. Using Cuboid Splatting-with-prior yields more accurate reconstructions than (i) the used dispersion simulations alone and (ii) plain Cuboid Splatting without prior. Across reconstructions, the mean scenario error is 19.61%, improving on (i) by 28.02pp and on (ii) by 89.86pp, the latter with particularly large gains at high spatial resolution. These results demonstrate that combining simulation-based priors with measurement-based posterior inference can substantially improve the reconstruction of iodine air activity concentrations in nuclear emergencies.
Spatio-temporal traffic forecasting is a core component of intelligent transportation systems, supporting various downstream tasks such as signal control and network-level traffic management. In real-world deployments, forecasting models must operate under structural and observational uncertainties, conditions that are rarely considered in model design. Recent approaches achieve strong short-term predictive performance by tightly coupling spatial and temporal modeling, often at the cost of increased complexity and limited modularity. In contrast, efficient time-series models capture long-range temporal dependencies without relying on explicit network structure. We propose UniST-Pred, a unified spatio-temporal forecasting framework that first decouples temporal modeling from spatial representation learning, then integrates both through adaptive representation-level fusion. To assess robustness of the proposed approach, we construct a dataset based on an agent-based, microscopic traffic simulator (MATSim) and evaluate UniST-Pred under severe network disconnection scenarios. Additionally, we benchmark UniST-Pred on standard traffic prediction datasets, demonstrating its competitive performance against existing well-established models despite a lightweight design. The results illustrate that UniST-Pred maintains strong predictive performance across both real-world and simulated datasets, while also yielding interpretable spatio-temporal representations under infrastructure disruptions. The source code and the generated dataset are available at https://anonymous.4open.science/r/UniST-Pred-EF27.
Accurate simulation of terrestrial ecological carbon cycles is crucial for global climate change and ecosystem management. Process-based carbon models have high interpretability, but suffer from insufficient accuracy and slow computation due to fixed parameters. In contrast, deep-learning carbon models achieve high accuracy, but disregard physical principles, which prevents ecologists from explaining ecosystem dynamics. We propose NeuroDALEC, an interpretable framework that embeds the DALEC carbon-cycle model within a neural network, enabling differentiable computation of ecological processes. Key parameters and ensemble learning strategies are designed, and mass-conserving carbon pool state transition equations are introduced to ensure physical consistency. Experiments show NeuroDALEC outperforms existing models in both accuracy and efficiency. Moreover, it provides sufficient interpretability by predicting all components of the carbon cycle. Deployed in a real-time carbon assimilation system, NeuroDALEC supports daily carbon forecasting and decision-making. This work contributes to the United Nations' Sustainable Development Goals 13 (Climate Action) and 15 (Life on Land). The source code is available at: https://github.com/codesiena/NeuroDALEC.
Classical poetry is central to aesthetic education and cultural inheritance in China's K-12 education, and web-based instructional videos have become a primary learning resource. However, existing production pipelines either require costly teacher-guided editing or generate videos automatically without systematic instructional design, leading to unclear logic, shallow explanations, and limited contextual adaptation. We propose PoemDirector, a multi-agent framework developed with a national K-12 educational platform partner that unifies context-adaptive instructional mode selection, hierarchical explanation, and end-to-end video generation. A Director Agent analyzes each poem, selects a pedagogically grounded instructional mode from a teacher-co-designed library, and coordinates specialized agents to produce a complete instructional video from a poem title alone. We further establish a multidimensional evaluation framework, refined by literary and education experts, to assess instructional quality and poetic presentation. Experiments show that PoemDirector substantially outperforms commercial baselines and approaches human-crafted video quality across multiple dimensions. The resources are publicly available at https://github.com/ai4ed/PoemDirector.
Reducing preventable maternal mortality remains a global health priority. Under Sustainable Development Goal (SDG) target 3.1, the WHO emphasizes timely and equitable allocation of limited maternal health resources. Motivated by Department of Obstetrics and Gynecology at several important hospitals in Uganda and Ghana, we study the problem of sequential allocation of wearable vital sign monitoring devices among maternal mothers. While deep reinforcement learning (RL) has shown promise for sequential resource allocation, its limited interpretability hinders adoption in such high-stakes settings. In contrast, large language model (LLM) agents provide human-readable reasoning but often struggle with effective long-term decision making. To bridge this gap, we introduce Rule-Bottleneck RL (RBRL), the first LLM agent framework for resource allocation problems that jointly optimizes language-based decision policy and explainability. At each step within RBRL, an LLM first generates candidate rules---language statements capturing decision priorities tailored to the current state. RL then optimizes rule selection to maximize environmental rewards and explainability, with the LLM acting as a judge. Finally, an LLM chooses the action (optimal allocation) based on the rule. We provide conditions for RBRL performance guarantees as well as the finite-horizon evaluation gap of the learned RBRL policy. Experiments in maternal health show that RBRL outperforms baseline LLM agents and approaches the performance of deep RL, while producing clearer, policy-relevant explanations. Human evaluations further confirm improved trust and usability, demonstrating RBRL as a practical AI approach aligned with SDG target 3.1.
Accurate mapping of agricultural land use activity, particularly long-term transition from cropland to pasture and short-term transition between cropland to fallow land, is essential for sustainable water management, drought response, and food-system resilience which directly supports United Nations Sustainable Development Goals (SDG-2 and SDG-8). However, reliable agricultural land use activity mapping is challenging due to spectral ambiguity, temporal irregularities, severe class imbalance, and limited generalization across agricultural regions. In this work, we propose AG-STELLA, a knowledge guided spatio-temporal model that (i) captures temporal changes of agricultural lands using pretrained spatio-temporal transformers; (ii) integrates geospatial context using AlphaEarth embedding; (iii) introduces a temporal transition latent space with temporal consistency constraints; (iv) employs guidance through hydroclimatic consistency; and (v) uses a land use-aware gated decoder to improve robustness across regions. Through experimentation across three water-stressed U.S. states, we show consistent gains over baseline vision and foundation models, achieving up to 27% F1-score improvement for pasture (minority class) and 16% overall. We further show the robustness across heterogeneous regions through cross-state transfer learning, where AG-STELLA consistently outperforms foundation model baselines and achieve up to 82.3% F1-score for fallow land with a 9.6% improvement over the best foundation model. Code: https://github.com/Nibir088/AG-STELLA
Online learning systems provide accessible learning opportunities for blind or low-vision students. To support access to complex scientific materials, the speech models used in these systems need to deliver accurate scientific formula verbalization. While recent large speech language models (LSLMs) provide remarkable low-latency streaming capabilities, their potential for scientific formula verbalization remains underexplored. In this paper, we propose Formula-Speech, the first end-to-end LSLM designed for scientific formula verbalization. Specifically, we construct two high-quality scientific formula datasets with educational experts to align speech models with scientific formula verbalization patterns. We then adopt a lightweight and effective two-stage training framework, combining supervised fine-tuning for basic formula-to-speech alignment with reinforcement learning guided by a custom reward function to optimize for human-preferred verbalization. Experimental results show that our model significantly improves the verbalization performance of LSLMs and achieves state-of-the-art results across multiple scientific domains. Our code and datasets are available at https://github.com/ai4ed/FormulaSpeech.
Credit card fraud threatens global payment ecosystems, causing billions in losses and undermining public trust. Efficient fraud detection remains challenging due to surging transaction volumes and evolving tactics. While Graph Neural Networks (GNNs) excel at modeling structural relationships, they struggle in real-world scenarios characterized by label scarcity and often overlook discriminative feature-level signals, leaving rich risk signals underutilized without costly manual engineering. To address this, we propose DRESS, a Deep Reinforcement Learning (DRL) Enhanced Semi-supervised GNN framework. It employs a DRL agent to automatically capture and enhance feature-level risks, fusing them with graph-based structural risks and propagating via a gated temporal attention network for final prediction. To mitigate inefficient exploration of the DRL module, we incorporate a feature self-attention layer to weigh feature contributions to fraud detection and employ self-supervised intrinsic rewards to help optimize the DRL module efficiently. Extensive experiments on real-world datasets demonstrate that DRESS outperforms state-of-the-art methods, especially in low-label scenarios with only 2%–10% labeled samples. By empowering resource-limited institutions to combat fraud and prevent financial loss, DRESS secures the digital trust essential for inclusive growth, contributing to AI for poverty alleviation and economic development.
Brazil is home to about 180 Indigenous languages, which span a wide range of sociolinguistic circumstances, from critically endangered to relatively stable. Across this continuum, Indigenous communities often lack pedagogical resources and are underserved by existing language-learning technologies, which are typically designed for high-resource languages and assume solid connectivity and large datasets. This research project proposes AI-assisted tools and workflows for the collection and annotation of textual and speech data that substantially reduce the time and cost required to produce engaging language-learning game apps. Our goal is to implement a language-learning game app that Indigenous students can use to practice their reading, writing and speaking skills at home. We propose novel speech processing models for low-resource Indigenous languages and offlline support in low-connectivity environments. Our project adopts a co-creation model that actively foster collaboration between Indigenous educators, linguists, and youth, is adapted to the their context, and complies with ethical guidelines. We outline an implementation plan with Bororo and Enawene Nawe communities to test our methods and, potentially, produce an AI-driven platform for Indigenous language education that is applicable across diverse sociolinguistic contexts in Brazil and beyond.
The increasing frequency and severity of climate-related disasters have intensified the need for real-time monitoring, early warning, and informed decision-making. Earth Observation (EO), powered by satellite data and Machine Learning (ML), offers powerful tools to meet these challenges. Foundation Models (FMs) have revolutionized EO ML by enabling general-purpose pretraining on large-scale remote sensing datasets. However most existing models rely on high-resolution satellite imagery with low revisit rates, limiting their suitability for fast-evolving phenomena and time-critical emergency response. In this work, we present HighFM, a first cut approach towards a FM for high-temporal-resolution, multispectral EO data. Leveraging over 2 TB of SEVIRI imagery from the Meteosat Second Generation (MSG) platform, we adapt the SatMAE masked autoencoding framework to learn robust spatiotemporal representations. To support real-time monitoring, we enhance the original architecture with fine-grained temporal encodings to capture short-term variability. The pretrained models are then fine-tuned on cloud masking and active fire detection tasks. We benchmark our SEVIRI-pretrained Vision Transformers against traditional baselines and recent geospatial FMs, demonstrating consistent gains across both balanced accuracy and IoU metrics. Our results highlight the potential of temporally dense geostationary data for real-time EO, offering a scalable path toward foundation models for disaster detection and tracking.
Effective flood risk management relies on accurate forecasting, yet the ”black box” nature of state-of-the-art Deep Learning models creates a barrier to trust and accountability in high-stakes public safety decisions. While existing Explainable AI (XAI) methods offer local attributions, they fail to provide the verifiable, operationally meaningful causal narratives required by disaster response authorities. To address this societal challenge, we propose Context-Aware Concept Distillation (CACD), a framework developed in collaboration with domain experts to distill opaque LSTMs into interpretable, hydrology-aware surrogate models. We introduce an unsupervised pipeline to discover a ”Hydrological Language” and a Residual Hypernetwork that dynamically modulates these concepts based on static basin characteristics. Evaluated on 5,203 basins globally, our model achieves high fidelity (Median NSE 0.70), significantly outperforming black-box baselines (e.g., Multi Layer Perceptrons) on unseen future data. By demonstrating that human-interpretable concepts are sufficient to reconstruct flood dynamics, this work balances AI accuracy with the transparency required for responsible environmental decision-making.
Codecfakes (CFs) are a type of speech deepfakes generated through Audio Language Models (ALMs), with Neural Audio Codecs (NACs) forming the core mechanism for speech encoding and generation. CFs exhibit distributional characteristics that differ from vocoder-based deepfakes, causing detectors trained on vocoder data to generalize poorly to CFs detection. Although this has led to the development of CF detection benchmarks, existing resources are largely confined to English—and to a limited extent Chinese—leaving South-East Asian (SEA) languages unexplored. To bridge this gap, we introduce SEA-CF, the first large-scale benchmark for CF detection spanning multiple SEA languages, diverse speaker profiles, and a wide range of NAC architectures. SEA-CF is constructed by synthesizing publicly available real speech corpora. Our experiments show that state-of-the-art (SOTA) CF detectors trained on English-centric datasets fail to generalize to SEA speech due to language-specific phonetic structures, tonal variations, and rich prosodic diversity. We further conduct a comprehensive zero-shot and fine-tuned evaluation of recent SOTA ALMs on SEA-CF. Fine-tuning the ALMs improves performance, however, these are very large being impractical for real-world application due to their scale, particularly in low-resource and latency-constrained settings. To address this limitation, we propose a novel small-ALM, GARUDA tailored for CF detection, which delivers strong performance while remaining lightweight. Extensive evaluations demonstrate that the proposed Small-ALM outperforms strong end-to-end and ALM-based baselines, establishing a new, practical direction for robust CF detection in SEA languages and beyond.
Automatically mapping and segmenting global mining footprints using remote sensing and deep learning is critical for monitoring the socio-environmental risks and impacts of mining, yet its progress is hindered by the scarcity of fine-grained annotated data. Although large-scale datasets with coarse boundaries are widely available, leveraging them to improve fine-grained segmentation is challenging due to significant domain shift. To address this, we propose MineC2FNet, a coarse-to-fine domain incremental learning framework that exploits abundant coarse data to enhance fine-grained mining footprint segmentation. MineC2FNet adopts a teacher–student architecture with attentive distillation at both the feature and prediction levels, selectively transferring generalized knowledge from the coarse domain while enabling boundary refinement using limited fine-grained data (fine domain). We further introduce an expertly validated dataset of 219 images with precise boundary annotations across diverse geographies and commodities. Extensive experiments against state-of-the-art approaches, including domain adaptation and domain incremental learning methods, demonstrate that MineC2FNet achieves superior performance while effectively handling domain shift. The dataset and code are publicly available at https://github.com/risqiutama/MineC2FNet.
Wildfires, especially extreme wildfires, can cause prolonged damage to natural environments, human lives and global economies. To mitigate these impacts, enhancing wildfire preparedness is critical. This research proposal introduces iFire AI, a collaborative project to develop a world-leading 3D immersive visualisation system that enables users to experience and understand extreme wildfire scenarios. iFire AI utilises our advanced 360-degree immersive system AVIE, which visualises interactive landscapes and wildfire events enhanced by AI techniques. Firstly, we propose a deep learning model for wildfire behaviour modelling that generates minute-resolution wildfire progression information for immersive visualisation, supported by a Sim2Real pipeline that integrates simulated and real-world data to address data insufficiency and enhance model development and evaluation. We also explore 3D tree reconstruction using 3D Gaussian splatting, creating visually realistic and computationally efficient tree models. iFire AI can enhance users' risk perception, situational awareness and collaborative decision-making, and thereby reduce risks due to extreme wildfires and promote sustainable development.
Inertial Confinement Fusion (ICF) holds transformative promise for sustainable, near-limitless clean energy, yet remains constrained by prohibitively high costs and limited experimental opportunities. This paper presents Human-in-the-Loop Meta Bayesian Optimization (HL-MBO), a framework that integrates expert knowledge with few-shot, uncertainty-aware machine learning to accelerate discovery in data-scarce, high-stakes scientific domains. HL-MBO introduces a meta-learned surrogate model with an expert-informed acquisition function to recommend candidate experiments. To foster trust and enable informed decisions, HL-MBO also provides interpretable explanations of its suggestions. We show that HL-MBO outperforms current BO methods for ICF energy yield optimization, as well as benchmarks in molecular optimization and critical-temperature maximization for superconducting materials.