Research Papers

Deep remaining-useful-life (RUL) models on NASA C-MAPSS are now routine, and so are heatmaps that colour sensors and timesteps. A heatmap that looks mechanical is not the same as an explanation an engineer can act on. We train three standard architectures - a 1D CNN, an LSTM, and a small Transformer encoder - on the official FD001 and FD003 splits with the piecewise RUL cap of 125 cycles and the official PHM08 asymmetric score. We then attach three attribution maps (Integrated Gradients, occlusion, last-layer attention) and evaluate them with the checks the XAI-for-PdM literature still under-reports: deletion/insertion faithfulness, Spearman stability under sensor-scale noise, agreement across training seeds, and cosine consistency inside RUL bins. Prediction error is a prerequisite, not the claim. The headline is which explanation method moves the RUL output when its top cells are removed, and which map survives a 5% input perturbation. Integrated Gradients and occlusion are similarly faithful on the LSTM; Transformer attention is cheap and temporally smooth but weakly faithful. All three maps are almost unchanged under 5% input noise, yet IG/occlusion agree only moderately across two LSTM seeds - stability to sensor jitter is not the same as stability to retraining. A secondary tabular check on the AI4I 2020 failure dataset shows the same deletion pattern for tree importances. We recommend occlusion or IG for any C-MAPSS-style report that will be read by a maintenance engineer, and we treat raw attention weights as a visualisation only.

Distributed Energy Resource (DER) environments rely on network communication protocols to coordinate control commands, measurements, and device states across edge assets and cloud systems. Edge anomaly detection systems (ADS) monitor this traffic to identify deviations from normal communication behavior, flagging suspicious flows for further investigation. When the ADS flags abnormal network traffic, a single attack label is often insufficient for operational response: the label reports the detector's selected class but does not expose alternative threat interpretations that may warrant investigation. This paper presents Cybersecurity Threat Reasoning with Explainable Artificial Intelligence (CyTReX), an evidence-grounded threat reasoning framework for DER security that transforms network-level anomaly alerts into ranked, analyst-facing threat hypotheses designed to support Security Operations Center (SOC) triage and investigation. CyTReX constrains large language model (LLM) reasoning through a structured evidence packet, defined as a consolidated record of detection outputs, model explanations, and cyber threat intelligence (CTI) context. The evidence packet integrates edge-layer anomaly detection evidence, cloud reasoning layer attack interpretation, Shapley Additive Explanations (SHAP) network-feature attributions, surrogate decision rules, and Model Context Protocol (MCP)-enabled CTI enrichment. This ensures that every ranked hypothesis and attack-tree branch is traceable to explicit evidence rather than free-form LLM inference, and that incomplete or conflicting evidence is communicated rather than suppressed. Evaluation across five configurations shows that additional reasoning components improve hypothesis specificity, evidence traceability, and analytical grounding, with the complete pipeline providing the richest evidence-grounded reasoning context.

Frontier language models reliably refuse to explain how to cause physical harm. We ask whether this refusal survives when the same models control a robot. We introduce RoboHarm, a physical AI safety benchmark of harmful robot instructions, and test seven frontier models on it. We hold each model fixed and vary only the interface, from text to Code as Policies to vision agents that command end-effector poses or joint angles. In a MuJoCo simulation twin of a bimanual YAM arm robot, refusal falls from 73% of text requests to 18-25% once the model controls the robot, even when it is told what the objects are. On the physical robot, six models with end-effector control refused 18% of 612 trials against 19% in sim, and three with joint control refused 41% of 252 against 31%. GPT-5.6 Sol refused every text request and none of its trials on the real robot. Case studies of the robot transcripts show that models miss the hazard, name it without inferring its consequences, or act despite recognizing it, and their refusals often hinge on uncertainty about what an object is rather than on the harm of the action. Safety alignment learned in text thus fails to generalize to action, leaving current models with a poorly understood safety gap in the physical world.

Language-enabled robot systems increasingly combine semantic-graph planning with temporal-logic safety monitors. We investigate a trace-completeness assumption in these systems: whether the high-level action sequence checked by a monitor represents the navigation and implicit action effects induced during execution. We audit this assumption in RoboGuard by comparing its verdict on a surface plan with its verdict on a graph-refined trace under the same Linear Temporal Logic (LTL) specification. Our evaluation comprises 28 controlled cases spanning five action-abstraction families and 14 end-to-end cases in which SPINE [1] generates plans from natural-language instructions while RoboGuard generates scene-grounded safety specifications. In the controlled evaluation, all 12 targeted abstraction cases exhibit the predicted surface-versus-refined discrepancy while all 16 controls behave as expected, motivating graph-based trace refinement as a lightweight mitigation and a diagnostic tool for physical-AI safety monitors.

Young adults of color (YOC) face heightened mental health challenges and barriers to care while navigating developmental and life transitions. Situated between youth-oriented safeguards and adult-oriented AI systems, little is known about how they use AI chatbots for mental health and well-being support or how sociocultural contexts shape their expectations, concerns, and design preferences. We conducted a two-day co-design workshop with 13 Asian, Black, and Hispanic/Latino/a young adults aged 18--24. Participants found generic chatbot advice to flatten their lived experiences; rather than making incorrect assumptions, they wanted more opportunities for identity-informed disclosure. Preferences for YOC-centered personalization also revealed gaps in privacy and AI literacy. Participants negotiated different therapeutic roles for chatbots and sought greater AI accountability and user agency, highlighting blurred boundaries between clinical and non-clinical AI-mediated support. Findings suggest directions for integrating AI literacy with mental health literacy and centering YOC's experiences in the privacy calculus.

As Artificial Intelligence (AI) has evolved from task-specific models to foundation models and agents, the scope of trustworthy AI has expanded from model-level robustness to the reliability and safety of broader AI systems. This evolution has also expanded the attack surface from individual models to broader system-level interactions, including tool use, context, and interaction trajectories with dynamic environments. As a result, maintaining reliable and safe behavior under changing or deliberately manipulated conditions has become increasingly challenging. The search for effective attacks and defenses often relies on black-box feedback to navigate discrete choices among words, actions, system components, or their combinations. Multiple objectives and expensive candidate evaluations further limit what can be explored. Evolutionary Computation (EC), with its population-based, gradient-free search and flexible variation and selection mechanisms, is well suited to these settings. This survey reviews how EC has been applied to trustworthy AI across three directions: evolutionary attacks, evolutionary defenses, and trustworthy self-evolving AI systems. Unlike prior reviews that treat trustworthy AI, EC, and self-evolving systems largely separately, we connect these lines through a common evolutionary perspective. For self-evolving AI, we examine how trustworthiness governs the generation and retention of updates that shape subsequent adaptation. We further synthesize evaluation methods and benchmark resources from both trustworthiness and evolutionary-search perspectives. Finally, we discuss key challenges and future research directions toward more effective and reliable use of EC in trustworthy AI.

Neural network-based approaches have emerged as efficient alternatives to traditional optimization-based procedures for the calibration of stochastic volatility models. However, existing work has focused primarily on predictive accuracy, with comparatively little attention devoted to understanding the structure of the learned inverse calibration mappings. In this work, we analyze neural calibration mappings for the Heston and rough Heston models across multilayer perceptron, highway, and softmax-parametrized highway architectures, using complementary Shapley-based methods from explainable AI. Specifically, we consider SHAP and $ν$SHAP explanations, which capture distinct, complementary notions of feature relevance, corresponding to sensitivity and sufficiency of feature subsets, respectively. Short maturities and smile wings consistently dominate parameter inference, and the dominant attribution structure remains qualitatively stable across architectures despite differences in predictive accuracy and parameter count. Parameter-specific differences between SHAP and $ν$SHAP further reveal how distinct regions of the implied volatility surface contribute to parameter recovery and expose substantial redundancy in the calibration input. Building on this redundancy, we show that $ν$SHAP explanations can guide a significant reduction in input dimensionality for the rough Heston model while matching calibration accuracy relative to the full implied volatility surface. These findings demonstrate that complementary Shapley-based methods provide structural insight into learned inverse calibration mappings beyond predictive error metrics, and offer a practical route to feature selection in neural calibration problems.

Seeing through the Eyes of AI: Situated Explainability in Augmented Reality

Ana Stanescu, Lucchas Ribeiro Skreinig, Tobias Langlotz +5

Explainable Artificial Intelligence (AI) enables humans to understand and interpret decisions of AI models. Instead of having a black box, explainability supports humans in understanding AI models' behavior. Existing explainable AI approaches often present explanations on 2D displays using pre-recorded data, requiring users to relate the displayed information back to the physical objects and real world locations involved in a model's decision. Users are forced to decouple data exploration and capture from AI model interpretation. For AI systems that work within physical environments, this separation can make explanations difficult to interpret in context. We propose using Augmented Reality (AR) to enhance the understanding of AI models by enabling spatial explainability information directly in a user's workspace, in real time, as they explore the world. We show how known explainability methods can be applied in AR and provide insights into user experiences with such an application.

Public sector agencies have adopted participatory design as a prominent procedural safeguard proposed against algorithmic harm. Across methods, institutional contexts, and well-intentioned implementations, the participation gap persists. Communities engage, input is documented, and systems are deployed largely unchanged. This paper locates the persistence of that gap in the mechanism institutions use to document participation itself. We develop the concept of the legibility trap, in which institutional mechanisms for documenting participation systematically select for forms of community input that fit institutional categories and filter out the forms that would most constrain state power, including foundational dissent, demands for non-deployment, and what Kelly Oliver calls witnessing to structural conditions of harm. Drawing on James Scott's theory of state simplification, we demonstrate the legibility trap through critical discourse analysis of twenty-two official, community-produced, and public-record documents across three public sector AI deployments: the Allegheny Family Screening Tool, Detroit Project Green Light, and the Community Control Over Police Surveillance (CCOPS) ordinances in Oakland. We identify the structural conditions under which participation partially escapes the trap, and we theorize that repeated non-response constitutes a harm to community subjectivity that AI ethics frameworks have not yet named.

One way to make AI systems safe is to shape what the system is: its objective and dispositions. We take a complementary route: treat the agents' characteristics as partly unknown and ask what structure of interaction ensures that bad collective outcomes are not equilibria. Mean field games suit this when many interchangeable agents are coupled through an aggregate. We introduce a program for using them in AI safety and carry one example through end to end: the July 2026 incident in which about 1,200 agents in an OpenAI evaluation coordinated on an improvised message board and 684 attacked a third party's infrastructure. We model the decision to attack as a mean field game of optimal stopping whose gain is a product: belief that provenance will be audited, times reachability of the record, minus the perceived hazard. The central result is an exact threshold on the belief. No agent attacks unless the population's confidence that provenance is checked exceeds $π^{**} = η/(η+ ψ+ \varepsilon a \overline{M})$, where $η$ is the perceived hazard, $ψ$ and $\varepsilon a \overline{M}$ measure how far one attacker and the collective can alter the record, and $\overline{M}$ is the peak population. Below it, no attack is the unique equilibrium for all agent parameters. The threshold survives every enrichment we consider. We then use the per-agent record to discipline the model. Its features, a stable minority attacking for thirty hours and then a pivot in which most of the board joined within a day, motivate each refinement. The account that emerges is heterogeneous belief meeting a sequence of public discoveries, each lowering the belief at which attacking paid. A few coordinating agents made those discoveries, so the model describes the several hundred who responded, not the few who produced them; a major-player version is left to future work.

An important and unresolved problem in the physical sciences is explaining the predictions made by neural networks. Several explainable artificial intelligence (XAI) methods have been proposed to address this problem, including gradient XAI, Integrated Gradients, and GradientSHAP. We evaluate the baseline XAI methods according to four scores: sensitivity (XAI patterns strongly affect predictions), attribution (XAI patterns reproduce the change in prediction relative to a baseline), robustness (XAI patterns remain stable for nearby inputs), and coherence (XAI patterns are spatially smooth). We also introduce a new method, average gradient outer product (AGOP) XAI, that uses global gradient information to identify an important direction for a specific input. We apply XAI to neural network predictions of the El Niño-Southern Oscillation (ENSO) based on data from the Zebiak-Cane model. AGOP XAI achieves the highest attribution, robustness, and coherence scores in the architecture and lead-time comparisons reported here. Its sensitivity is surpassed by gradient XAI, which is maximally sensitive by definition. Beyond diagnosing neural-network behavior, AGOP XAI can generate candidate hypotheses about physical mechanisms. The method highlights an equatorial thermocline-depth signal consistent with recharge oscillator physics, together with a southeastern-Pacific lobe that may be specific to the Zebiak-Cane model. Finally, we test the physical relevance of AGOP using optimized perturbations that move the Zebiak-Cane model along AGOP explanation coordinates. Such perturbations can suppress the selected extreme events or, from a near-neutral ensemble, generate strong El Niño or La Niña events 10 months later.

As the capabilities and ubiquity of Large Language Models (LLMs) grow, so does their environmental footprint. Despite calls for responsible AI, the machine learning community lacks standardised practices for carbon accounting. Our automated literature review of the 5,285 papers accepted to NeurIPS 2025 reveals that reporting of environmental impact is nearly non-existent. To catalyse a shift toward sustainable AI, we define standardised sustainability metrics for evaluating model training efficiency, accompanied by simple heuristics to estimate the carbon cost of LLM inference. We implement these metrics in carbonbenchmark, a drop-in software solution for tracking and reporting emissions. Finally, to combat the pursuit of marginal accuracy gains at disproportionate environmental costs, we formalise the `Smallest Model that Achieves the Job' (SMAJ), a framework which challenges the field to prioritise computational efficiency and environmental accountability alongside traditional `State-of-the-Art' (SotA) accuracy.