Research Papers

Understanding and analyzing video actions are essential for producing insightful and contextualized descriptions, especially for video-based applications like intelligent monitoring and autonomous systems. The proposed work introduces a novel framework for generating natural language descriptions from video datasets by combining textual and visual modalities. The suggested architecture makes use of ResNet50 to extract visual features from video frames that are taken from the Microsoft Research Video Description Corpus (MSVD), and Berkeley DeepDrive eXplanation (BDD-X) datasets. The extracted visual characteristics are converted into patch embeddings and then run through an encoder-decoder model based on Generative Pre-trained Transformer-2 (GPT-2). In order to align textual and visual representations and guarantee high-quality description production, the system uses multi-head self-attention and cross-attention techniques. The model's efficacy is demonstrated by performance evaluation using BLEU (1-4), CIDEr, METEOR, and ROUGE-L. The suggested framework outperforms traditional methods with BLEU-4 scores of 0.755 (BDD-X) and 0.778 (MSVD), CIDEr scores of 1.235 (BDD-X) and 1.315 (MSVD), METEOR scores of 0.312 (BDD-X) and 0.329 (MSVD), and ROUGE-L scores of 0.782 (BDD-X) and 0.795 (MSVD). By producing human-like, contextually relevant descriptions, strengthening interpretability, and improving real-world applications, this research advances explainable AI.

Due to the proliferation of short-form content and the rapid adoption of AI, opportunities for deep, reflective thinking have significantly diminished, undermining users' critical thinking and reducing engagement with the reasoning behind AI-generated outputs. To address this issue, we propose an Interactive Chain-of-Thought (CoT) Framework that enhances human-centered explainability and responsible AI usage by making the model's inference process transparent, modular, and user-editable. The framework decomposes reasoning into clearly defined blocks that users can inspect, modify, and re-execute, encouraging active cognitive engagement rather than passive consumption. It further integrates a lightweight edit-adaptation mechanism inspired by preference learning, allowing the system to align with diverse cognitive styles and user intentions. Ethical transparency is ensured through explicit metadata disclosure, built-in bias checkpoint functionality, and privacy-preserving safeguards. This work outlines the design principles and architecture necessary to promote critical engagement, responsible interaction, and inclusive adaptation in AI systems aimed at addressing complex societal challenges.

Machine learning models tend to learn spurious features - features that strongly correlate with target labels but are not causal. Existing approaches to mitigate models' dependence on spurious features work in some cases, but fail in others. In this paper, we systematically analyze how and where neural networks encode spurious correlations. We introduce the neuron spurious score, an XAI-based diagnostic measure to quantify a neuron's dependence on spurious features. We analyze both convolutional neural networks (CNNs) and vision transformers (ViTs) using architecture-specific methods. Our results show that spurious features are partially disentangled, but the degree of disentanglement varies across model architectures. Furthermore, we find that the assumptions behind existing mitigation methods are incomplete. Our results lay the groundwork for the development of novel methods to mitigate spurious correlations and make AI models safer to use in practice.

Machine learning methods find growing application in the reconstruction and analysis of data in high energy physics experiments. A modified convolutional autoencoder model was employed to identify and reconstruct the pulses from scintillating crystals. The model was further investigated using four xAI methods for deeper understanding of the underlying reconstruction mechanism. The results are discussed in detail, underlining the importance of xAI for knowledge gain and further improvement of the algorithms.

The issue of fairness in decision-making is a critical one, especially given the variety of stakeholder demands for differing and mutually incompatible versions of fairness. Adopting a strategic interaction of perspectives provides an alternative to enforcing a singular standard of fairness. We present a web-based software application, FairPlay, that enables multiple stakeholders to debias datasets collaboratively. With FairPlay, users can negotiate and arrive at a mutually acceptable outcome without a universally agreed-upon theory of fairness. In the absence of such a tool, reaching a consensus would be highly challenging due to the lack of a systematic negotiation process and the inability to modify and observe changes. We have conducted user studies that demonstrate the success of FairPlay, as users could reach a consensus within about five rounds of gameplay, illustrating the application's potential for enhancing fairness in AI systems.

Risk thresholds provide a measure of the level of risk exposure that a society or individual is willing to withstand, ultimately shaping how we determine the safety of technological systems. Against the backdrop of the Cold War, the first risk analyses, such as those devised for nuclear systems, cemented societally accepted risk thresholds against which safety-critical and defense systems are now evaluated. But today, the appropriate risk tolerances for AI systems have yet to be agreed on by global governing efforts, despite the need for democratic deliberation regarding the acceptable levels of harm to human life. Absent such AI risk thresholds, AI technologists-primarily industry labs, as well as "AI safety" focused organizations-have instead advocated for risk tolerances skewed by a purported AI arms race and speculative "existential" risks, taking over the arbitration of risk determinations with life-or-death consequences, subverting democratic processes. In this paper, we demonstrate how such approaches have allowed AI technologists to engage in "safety revisionism," substituting traditional safety methods and terminology with ill-defined alternatives that vie for the accelerated adoption of military AI uses at the cost of lowered safety and security thresholds. We explore how the current trajectory for AI risk determination and evaluation for foundation model use within national security is poised for a race to the bottom, to the detriment of the US's national security interests. Safety-critical and defense systems must comply with assurance frameworks that are aligned with established risk thresholds, and foundation models are no exception. As such, development of evaluation frameworks for AI-based military systems must preserve the safety and security of US critical and defense infrastructure, and remain in alignment with international humanitarian law.

The Web3 ecosystem, underpinned by cryptographic primitives and decentralized consensus, represents a high-stakes environment where software vulnerabilities and incentive misalignments translate directly into financial loss. As Large Language Models (LLMs) are increasingly integrated into this domain for tasks ranging from smart contract auditing to decentralized finance analytics, ensuring their reliability is paramount. However, general-purpose benchmarks fail to capture the specialized reasoning required for these adversarial and protocol-driven settings. To bridge this gap, we introduce DMind Benchmark, a comprehensive evaluation suite designed to rigorously assess LLM proficiency across the Web3 stack. DMind Benchmark encompasses nine distinct subdomains (spanning infrastructure, smart contracts, token economics, etc.) and combines objective knowledge retrieval with complex open-ended reasoning tasks that emulate real-world operational challenges. We conduct an extensive evaluation of 31 leading proprietary and open-weights models, employing a contamination-aware pipeline and verifying the statistical robustness of our scoring protocol through rigorous cross-judge consistency checks. Our analysis reveals a critical dichotomy: while models demonstrate competence in foundational infrastructure concepts, they exhibit significant vulnerabilities in high-reasoning tasks such as security auditing. Furthermore, we provide a Pareto analysis to guide cost-effective deployment and demonstrate through adversarial experiments that high performance on DMind Benchmark necessitates genuine reasoning rather than superficial memorization. Since its open-source release in April 2025, DMind Benchmark achieved the #1 trending position on Hugging Face for nearly a week and accumulated over 13k downloads by June 2026, establishing itself as a standard for advancing secure and trustworthy AI in Web3.

Explanations for artificial intelligence (AI) systems are intended to support the people who are impacted by AI systems in high-stakes decision-making environments, such as doctors, patients, teachers, students, housing applicants, and many others. To protect people and support the responsible development of AI, explanations need to be actionable--helping people take pragmatic action in response to an AI system--and contestable--enabling people to push back against an AI system and its determinations. For many high-stakes domains, such as healthcare, education, and finance, the sociotechnical environment includes significant legal implications that impact how people use AI explanations. For example, physicians who use AI decision support systems may need information on how accepting or rejecting an AI determination will protect them from lawsuits or help them advocate for their patients. In this paper, we make the case for Legally-Informed Explainable AI, responding to the need to integrate and design for legal considerations when creating AI explanations. We describe three stakeholder groups with different informational and actionability needs, and provide practical recommendations to tackle design challenges around the design of explainable AI systems that incorporate legal considerations.

LLM jailbreaks are a widespread safety challenge. Given this problem has not yet been tractable, we suggest targeting a key failure mechanism: the failure of safety to generalize across semantically equivalent inputs. We further focus the target by requiring desirable tractability properties of attacks to study: explainability, transferability between models, and transferability between goals. We perform red-teaming within this framework by uncovering new vulnerabilities to multi-turn, multi-image, and translation-based attacks. These attacks are semantically equivalent by our design to their single-turn, single-image, or untranslated counterparts, enabling systematic comparisons; we show that the different structures yield different safety outcomes. We then demonstrate the potential for this framework to enable new defenses by proposing a Structure Rewriting Guardrail, which converts an input to a structure more conducive to safety assessment. This guardrail significantly improves refusal of harmful inputs, without over-refusing benign ones. Thus, by framing this intermediate challenge - more tractable than universal defenses but essential for long-term safety - we highlight a critical milestone for AI safety research.

The renaissance of deep learning has led to the massive development of automated driving. However, deep neural networks are vulnerable to adversarial examples. The perturbations of adversarial examples are imperceptible to human eyes but can lead to the false predictions of neural networks. It poses a huge risk to artificial intelligence (AI) applications for automated driving. This survey systematically reviews the development of adversarial robustness research over the past decade, including the attack and defense methods and their applications in automated driving. The growth of automated driving pushes forward the realization of trustworthy AI applications. This review lists significant references in the research history of adversarial examples.

Mild Cognitive Impairment (MCI) affects 12-18% of individuals over 60. MCI patients exhibit cognitive dysfunctions without significant daily functional loss. While MCI may progress to dementia, predicting this transition remains a clinical challenge due to limited and unreliable indicators. Behavioral changes, like in the execution of Activities of Daily Living (ADLs), can signal such progression. Sensorized smart homes and wearable devices offer an innovative solution for continuous, non-intrusive monitoring ADLs for MCI patients. However, current machine learning models for detecting behavioral changes lack transparency, hindering clinicians' trust. This paper introduces the SERENADE project, a European Union-funded initiative that aims to detect and explain behavioral changes associated with cognitive decline using explainable AI methods. SERENADE aims at collecting one year of data from 30 MCI patients living alone, leveraging AI to support clinical decision-making and offering a new approach to early dementia detection.

Focal cortical dysplasia (FCD) type II is a major cause of drug-resistant epilepsy, often curable only by surgery. Despite its clinical importance, the diagnosis of FCD is very difficult in MRI because of subtle abnormalities, leading to misdiagnosis. This study investigates the use of 3D convolutional neural networks (3D-CNNs) for FCD detection, using a dataset of 170 subjects (85 FCD patients and 85 controls) composed of T1-weighted and FLAIR MRI scans. In particular, it investigates the benefits obtained from cross-modality transfer learning and explainable artificial intelligence (XAI) techniques, in particular Gradient-weighted Class Activation Mapping (Grad-CAM). ResNet architectures (ResNet-18, -34, and -50) were implemented, employing transfer learning strategies that used pre-trained weights from segmentation tasks. Results indicate that transfer learning significantly enhances classification accuracy (up to 80.3%) and interpretability, as measured by a novel Heat-Score metric, which evaluates the model's focus on clinically relevant regions. Improvements in the Heat-Score metric underscore the model's seizure zone localization capabilities, bringing AI predictions and clinical insights closer together. These results highlight the importance of transfer learning, including cross-modality, and XAI in advancing AI-based medical diagnostics, especially for difficult-to-diagnose pathologies such as FCD.