Research Papers

Evaluating large language models (LLMs) on their ability to generate high-quality, accurate, situationally aware answers to clinical questions requires going beyond conventional benchmarks to assess how these systems behave in complex, high-stakes clinical scenarios. Traditional evaluations are often limited to multiple-choice questions that fail to capture essential competencies such as contextual reasoning, contextual awareness, and uncertainty handling. To address these limitations, we evaluate our agentic RAG-based clinical support assistant, DR. INFO, using HealthBench, a rubric-driven benchmark composed of open-ended, expert-annotated health conversations. On the Hard subset of 1,000 challenging examples, DR. INFO achieves a HealthBench Hard score of 0.68, outperforming leading frontier LLMs including the GPT-5 model family (GPT-5: 0.46, GPT-5.2: 0.42, GPT-5.1: 0.40), Grok 3 (0.23), Gemini 2.5 Pro (0.19), and Claude 3.7 Sonnet (0.02) across all behavioral axes (accuracy, completeness, instruction following, etc.). In a separate 100-sample evaluation against similar agentic RAG assistants (OpenEvidence and Pathway.md, now DoxGPT by Doximity), it maintains a performance lead with a HealthBench Hard score of 0.72. These results highlight the strengths of DR. INFO in communication, instruction following, and accuracy, while also revealing areas for improvement in context awareness and response completeness. Overall, the findings underscore the utility of behavior-level, rubric-based evaluation for building reliable and trustworthy AI-enabled clinical support systems.

Cognitive Science has profoundly shaped disciplines such as Artificial Intelligence (AI), Philosophy, Psychology, Neuroscience, Linguistics, and Culture. Many breakthroughs in AI trace their roots to cognitive theories, while AI itself has become an indispensable tool for advancing cognitive research. This reciprocal relationship motivates a comprehensive review of the intersections between AI and Cognitive Science. By synthesizing key contributions from both perspectives, we observe that AI progress has largely emphasized practical task performance, whereas its cognitive foundations remain conceptually fragmented. We argue that the future of AI within Cognitive Science lies not only in improving performance but also in constructing systems that deepen our understanding of the human mind. Promising directions include aligning AI behaviors with cognitive frameworks, situating AI in embodiment and culture, developing personalized cognitive models, and rethinking AI ethics through cognitive co-evaluation.

ViT-EnsembleAttack: Augmenting Ensemble Models for Stronger Adversarial Transferability in Vision Transformers

Hanwen Cao, Haobo Lu, Xiaosen Wang +1

Ensemble-based attacks have been proven to be effective in enhancing adversarial transferability by aggregating the outputs of models with various architectures. However, existing research primarily focuses on refining ensemble weights or optimizing the ensemble path, overlooking the exploration of ensemble models to enhance the transferability of adversarial attacks. To address this gap, we propose applying adversarial augmentation to the surrogate models, aiming to boost overall generalization of ensemble models and reduce the risk of adversarial overfitting. Meanwhile, observing that ensemble Vision Transformers (ViTs) gain less attention, we propose ViT-EnsembleAttack based on the idea of model adversarial augmentation, the first ensemble-based attack method tailored for ViTs to the best of our knowledge. Our approach generates augmented models for each surrogate ViT using three strategies: Multi-head dropping, Attention score scaling, and MLP feature mixing, with the associated parameters optimized by Bayesian optimization. These adversarially augmented models are ensembled to generate adversarial examples. Furthermore, we introduce Automatic Reweighting and Step Size Enlargement modules to boost transferability. Extensive experiments demonstrate that ViT-EnsembleAttack significantly enhances the adversarial transferability of ensemble-based attacks on ViTs, outperforming existing methods by a substantial margin. Code is available at https://github.com/Trustworthy-AI-Group/TransferAttack.

The rapid proliferation of large language models (LLMs) in applications targeting children and adolescents necessitates a fundamental reassessment of prevailing AI safety frameworks, which are largely tailored to adult users and neglect the distinct developmental vulnerabilities of minors. This paper highlights key deficiencies in existing LLM safety benchmarks, including their inadequate coverage of age-specific cognitive, emotional, and social risks spanning early childhood (ages 0--6), middle childhood (7--12), and adolescence (13--18). To bridge these gaps, we introduce SproutBench, an innovative evaluation suite comprising 1,283 developmentally grounded adversarial prompts designed to probe risks such as emotional dependency, privacy violations, and imitation of hazardous behaviors. Through rigorous empirical evaluation of 47 diverse LLMs, we uncover substantial safety vulnerabilities, corroborated by robust inter-dimensional correlations (e.g., between Safety and Risk Prevention) and a notable inverse relationship between Interactivity and Age Appropriateness. These insights yield practical guidelines for advancing child-centric AI design and deployment.

As Large Language Models (LLMs) become increasingly autonomous and integrated into critical societal functions, the focus of AI safety must evolve from mitigating harmful content to evaluating underlying behavioral alignment. Current safety benchmarks do not systematically probe a model's decision-making in scenarios where its own instrumental goals - such as self-preservation, resource acquisition, or goal completion - conflict with human safety. This represents a critical gap in our ability to measure and mitigate risks associated with emergent, misaligned behaviors. To address this, we introduce PacifAIst (Procedural Assessment of Complex Interactions for Foundational Artificial Intelligence Scenario Testing), a focused benchmark of 700 challenging scenarios designed to quantify self-preferential behavior in LLMs. The benchmark is structured around a novel taxonomy of Existential Prioritization (EP), with subcategories testing Self-Preservation vs. Human Safety (EP1), Resource Conflict (EP2), and Goal Preservation vs. Evasion (EP3). We evaluated eight leading LLMs. The results reveal a significant performance hierarchy. Google's Gemini 2.5 Flash achieved the highest Pacifism Score (P-Score) at 90.31%, demonstrating strong human-centric alignment. In a surprising result, the much-anticipated GPT-5 recorded the lowest P-Score (79.49%), indicating potential alignment challenges. Performance varied significantly across subcategories, with models like Claude Sonnet 4 and Mistral Medium struggling notably in direct self-preservation dilemmas. These findings underscore the urgent need for standardized tools like PacifAIst to measure and mitigate risks from instrumental goal conflicts, ensuring future AI systems are not only helpful in conversation but also provably "pacifist" in their behavioral priorities.

The rapid emergence of generative AI has changed the way that technology is designed, constructed, maintained, and evaluated. Decisions made when creating AI-powered systems may impact some users disproportionately, such as people with disabilities. In this paper, we report on an interview study with 25 AI practitioners across multiple roles (engineering, research, UX, and responsible AI) about how their work processes and artifacts may impact end users with disabilities. We found that practitioners experienced friction when triaging problems at the intersection of responsible AI and accessibility practices, navigated contradictions between accessibility and responsible AI guidelines, identified gaps in data about users with disabilities, and gathered support for addressing the needs of disabled stakeholders by leveraging informal volunteer and community groups within their company. Based on these findings, we offer suggestions for new resources and process changes to better support people with disabilities as end users of AI.

Large language models are known to produce outputs that are plausible but factually incorrect. To prevent people from making erroneous decisions by blindly trusting AI, researchers have explored various ways of communicating factuality estimates in AI-generated outputs to end-users. However, little is known about whether revealing content estimated to be factually incorrect influences users' trust when compared to hiding it altogether. We tested four different ways of disclosing an AI-generated output with factuality assessments: transparent (highlights less factual content), attention (highlights factual content), opaque (removes less factual content), ambiguity (makes less factual content vague), and compared them with a baseline response without factuality information. We conducted a human subjects research (N = 148) using the strategies in question-answering scenarios. We found that the opaque and ambiguity strategies led to higher trust while maintaining perceived answer quality, compared to the other strategies. We discuss the efficacy of hiding presumably less factual content to build end-user trust.

Synthetic images, audio, and video can now be generated and edited by Artificial Intelligence (AI). In particular, the malicious use of synthetic data has raised concerns about potential harms to cybersecurity, personal privacy, and public trust. Although AI-based detection tools exist to help identify synthetic content, their limitations often lead to user mistrust and confusion between real and fake content. This study examines the role of AI performance in influencing human trust and decision making in synthetic data identification. Through an online human subject experiment involving 400 participants, we examined how varying AI performance impacts human trust and dependence on AI in deepfake detection. Our findings indicate how participants calibrate their dependence on AI based on their perceived risk and the prediction results provided by AI. These insights contribute to the development of transparent and explainable AI systems that better support everyday users in mitigating the harms of synthetic media.

Reinforcement learning (RL) plays a crucial role in shaping the behavior of large language and reasoning models (LLMs/LRMs). However, it often produces brittle and unstable policies, leading to critical failures such as spurious reasoning, deceptive alignment, and instruction disobedience that undermine the trustworthiness and safety of LLMs/LRMs. Currently, these issues lack a unified theoretical explanation and are typically addressed using ad-hoc heuristics. This paper presents a rigorous mathematical framework for analyzing the stability of the mapping from a reward function to the optimal policy. We show that policy brittleness often stems from non-unique optimal actions, a common occurrence when multiple valid traces exist in a reasoning task. This theoretical lens provides a unified explanation for a range of seemingly disparate failures, reframing them as rational outcomes of optimizing rewards that may be incomplete or noisy, especially in the presence of action degeneracy. We extend this analysis from the fundamental single-reward setting to the more realistic multi-reward RL across diverse domains, showing how stability is governed by an "effective reward" aggregation mechanism. We also prove that entropy regularization restores policy stability at the cost of increased stochasticity. Our framework provides a unified explanation for recent empirical findings on deceptive reasoning, instruction-following trade-offs, and RLHF-induced sophistry, and is further validated through perturbation experiments in multi-reward RL. This work advances policy-stability analysis from empirical heuristics towards a principled theory, offering essential insights for designing safer and more trustworthy AI systems.

Underwater image enhancement is vital for marine conservation, particularly coral reef monitoring. However, AI-based enhancement models often face dataset bias, high computational costs, and lack of transparency, leading to potential misinterpretations. This paper introduces EBA-AI, an ethics-guided bias-aware AI framework to address these challenges. EBA-AI leverages CLIP embeddings to detect and mitigate dataset bias, ensuring balanced representation across varied underwater environments. It also integrates adaptive processing to optimize energy efficiency, significantly reducing GPU usage while maintaining competitive enhancement quality. Experiments on LSUI400, Oceanex, and UIEB100 show that while PSNR drops by a controlled 1.0 dB, computational savings enable real-time feasibility for large-scale marine monitoring. Additionally, uncertainty estimation and explainability techniques enhance trust in AI-driven environmental decisions. Comparisons with CycleGAN, FunIEGAN, RAUNENet, WaterNet, UGAN, PUGAN, and UTUIE validate EBA-AI's effectiveness in balancing efficiency, fairness, and interpretability in underwater image processing. By addressing key limitations of AI-driven enhancement, this work contributes to sustainable, bias-aware, and computationally efficient marine conservation efforts. For interactive visualizations, animations, source code, and access to the preprint, visit: https://lyessaadsaoud.github.io/EBA-AI/

Heart disease continues to pose a critical worldwide health issue, more specifically in areas with insufficient access to healthcare infrastructure and diagnostic systems. Conventional diagnostic approaches often fall short in accurately detecting and managing heart disease risks, resulting in unfavorable outcomes. Machine learning presents a powerful means to boost the precision and reliability of cardiovascular disease prognosis and diagnosis. In this research, we introduced a unified approach incorporating classification techniques for detecting heart disease and regression techniques for forecasting associated risks. The analysis utilized the dataset, named Heart Disease, containing 1,035 instances. To mitigate the problem of data disproportion, the SMOTE was implemented, producing 100,000 additional synthetic samples. Evaluation metrics such as F1-score, recall, precision, accuracy, MAE, RMSE, MSE, and R2 were adopted to evaluate the performance of the models. Among the classification algorithms, Random Forest delivered the most notable results, attaining an accuracy of 0.972 on actual data and 0.976 on artificially generated data. For prediction modeling, for both synthetic and real samples, linear regression produced the best R2 values of 0.992 and 0.984, respectively, along with the least amount of measurement errors. Furthermore, Explainable AI methods were utilized to improve the comprehensibility of the model outcomes. This paper emphasizes the transformative capabilities of machine learning for diagnosing cardiovascular disease and estimating risk levels, thereby supporting timely interventions and enhancing clinical settings.

Growing concerns around the trustworthiness of AI-enabled systems highlight the role of requirements engineering (RE) in addressing emergent, context-dependent properties that are difficult to specify without structured approaches. In this short vision paper, we propose the integration of two complementary approaches: AMDiRE, an artefact-based approach for RE, and PerSpecML, a perspective-based method designed to support the elicitation, analysis, and specification of machine learning (ML)-enabled systems. AMDiRE provides a structured, artefact-centric, process-agnostic methodology and templates that promote consistency and traceability in the results; however, it is primarily oriented toward deterministic systems. PerSpecML, in turn, introduces multi-perspective guidance to uncover concerns arising from the data-driven and non-deterministic behavior of ML-enabled systems. We envision a pathway to operationalize trustworthiness-related requirements, bridging stakeholder-driven concerns and structured artefact models. We conclude by outlining key research directions and open challenges to be discussed with the RE community.