Vision-language models (VLMs) enable open-ended visual question answering but remain prone to hallucinations. We present HEDGE, a unified framework for hallucination detection that combines controlled visual perturbations, semantic clustering, and robust uncertainty metrics. HEDGE integrates sampling, distortion synthesis, clustering (entailment- and embedding-based), and metric computation into a reproducible pipeline applicable across multimodal architectures. Evaluations on VQA-RAD and KvasirVQA-x1 with three representative VLMs (LLaVA-Med, Med-Gemma, Qwen2.5-VL) reveal clear architecture- and prompt-dependent trends. Hallucination detectability is highest for unified-fusion models with dense visual tokenization (Qwen2.5-VL) and lowest for architectures with restricted tokenization (Med-Gemma). Embedding-based clustering often yields stronger separation when applied directly to the generated answers, whereas NLI-based clustering remains advantageous for LLaVA-Med and for longer, sentence-level responses. Across configurations, the VASE metric consistently provides the most robust hallucination signal, especially when paired with embedding clustering and a moderate sampling budget (n ~ 10-15). Prompt design also matters: concise, label-style outputs offer clearer semantic structure than syntactically constrained one-sentence responses. By framing hallucination detection as a geometric robustness problem shaped jointly by sampling scale, prompt structure, model architecture, and clustering strategy, HEDGE provides a principled, compute-aware foundation for evaluating multimodal reliability. The hedge-bench PyPI library enables reproducible and extensible benchmarking, with full code and experimental resources available at https://github.com/Simula/HEDGE .

Foundation models for medical image segmentation struggle under out-of-distribution (OOD) shifts, often producing fragmented false positives on OOD tumors. We introduce R$^{2}$Seg, a training-free framework for robust OOD tumor segmentation that operates via a two-stage Reason-and-Reject process. First, the Reason step employs an LLM-guided anatomical reasoning planner to localize organ anchors and generate multi-scale ROIs. Second, the Reject step applies two-sample statistical testing to candidates generated by a frozen foundation model (BiomedParse) within these ROIs. This statistical rejection filter retains only candidates significantly different from normal tissue, effectively suppressing false positives. Our framework requires no parameter updates, making it compatible with zero-update test-time augmentation and avoiding catastrophic forgetting. On multi-center and multi-modal tumor segmentation benchmarks, R$^{2}$Seg substantially improves Dice, specificity, and sensitivity over strong baselines and the original foundation models. Code are available at https://github.com/Eurekashen/R2Seg.

Deploying embodied agents that can answer questions about their surroundings in realistic real-world settings remains difficult, partly due to the scarcity of benchmarks that faithfully capture practical operating conditions. We propose infrastructure inspection as a compelling domain for open-vocabulary Embodied Question Answering (EQA): it naturally demands multi-scale reasoning, long-range spatial understanding, and complex semantic relationships, while offering unique evaluation advantages via standardized National Bridge Inventory (NBI) condition ratings (0-9), professional inspection reports, and egocentric imagery. We introduce BridgeEQA, a benchmark of 2,200 open-vocabulary question-answer pairs (in the style of OpenEQA) grounded in professional inspection reports across 200 real-world bridge scenes with 47.93 images on average per scene. Questions require synthesizing visual evidence across multiple images and aligning responses with NBI condition ratings. We further propose a new EQA metric Image Citation Relevance to evaluate the ability of a model to cite relevant images. Evaluations of state-of-the-art vision-language models reveal substantial performance gaps under episodic memory EQA settings. To address this, we propose Embodied Memory Visual Reasoning (EMVR), which formulates inspection as sequential navigation over an image-based scene graph: images are nodes, and an agent takes actions to traverse views, compare evidence, and reason within a Markov decision process. EMVR shows strong performance over the baselines. We publicly release both the dataset and code.

Assurance for artificial intelligence (AI) systems remains fragmented across software supply-chain security, adversarial machine learning, and governance documentation. Existing transparency mechanisms - including Model Cards, Datasheets, and Software Bills of Materials (SBOMs) - advance provenance reporting but rarely provide verifiable, machine-readable evidence of model security. This paper introduces the AI Risk Scanning (AIRS) Framework, a threat-model-based, evidence-generating framework designed to operationalize AI assurance. The AIRS Framework evolved through three progressive pilot studies - Smurf (AIBOM schema design), OPAL (operational validation), and Pilot C (AIRS) - that reframed AI documentation from descriptive disclosure toward measurable, evidence-bound verification. The framework aligns its assurance fields to the MITRE ATLAS adversarial ML taxonomy and automatically produces structured artifacts capturing model integrity, packaging and serialization safety, structural adapters, and runtime behaviors. Currently, the AIRS Framework is scoped to provide model-level assurances for LLMs, but it could be expanded to include other modalities and cover system-level threats (e.g. application-layer abuses, tool-calling). A proof-of-concept on a quantized GPT-OSS-20B model demonstrates enforcement of safe loader policies, per-shard hash verification, and contamination and backdoor probes executed under controlled runtime conditions. Comparative analysis with SBOM standards of SPDX 3.0 and CycloneDX 1.6 reveals alignment on identity and evaluation metadata, but identifies critical gaps in representing AI-specific assurance fields. The AIRS Framework thus extends SBOM practice to the AI domain by coupling threat modeling with automated, auditable evidence generation, providing a principled foundation for standardized, trustworthy, and machine-verifiable AI risk documentation.

Federated learning (FL) enables multiple clients to collaboratively train a shared global model while preserving the privacy of their local data. Within this paradigm, the intellectual property rights (IPR) of client models are critical assets that must be protected. In practice, the central server responsible for maintaining the global model may maliciously manipulate the global model to erase client contributions or falsely claim sole ownership, thereby infringing on clients' IPR. Watermarking has emerged as a promising technique for asserting model ownership and protecting intellectual property. However, existing FL watermarking approaches remain limited, suffering from potential watermark collisions among clients, insufficient watermark security, and non-intuitive verification mechanisms. In this paper, we propose FLClear, a novel framework that simultaneously achieves collision-free watermark aggregation, enhanced watermark security, and visually interpretable ownership verification. Specifically, FLClear introduces a transposed model jointly optimized with contrastive learning to integrate the watermarking and main task objectives. During verification, the watermark is reconstructed from the transposed model and evaluated through both visual inspection and structural similarity metrics, enabling intuitive and quantitative ownership verification. Comprehensive experiments conducted over various datasets, aggregation schemes, and attack scenarios demonstrate the effectiveness of FLClear and confirm that it consistently outperforms state-of-the-art FL watermarking methods.

Aligning Large Language Models (LLMs) to be faithful to new knowledge in complex, multi-hop reasoning tasks is a critical, yet unsolved, challenge. We find that SFT-based methods, e.g., Reason-KE, while state-of-the-art, suffer from a "faithfulness gap": they optimize for format mimicry rather than sound reasoning. This gap enables the LLM's powerful parametric priors to override new contextual facts, resulting in critical factual hallucinations (e.g., incorrectly reasoning "Houston" from "NASA" despite an explicit edit). To solve this core LLM alignment problem, we propose Reason-KE++, an SFT+RL framework that instills process-level faithfulness. Its core is a Stage-aware Reward mechanism that provides dense supervision for intermediate reasoning steps (e.g., Decomposition, Sub-answer Correctness). Crucially, we identify that naive outcome-only RL is a deceptive trap for LLM alignment: it collapses reasoning integrity (e.g., 19.00% Hop acc) while superficially boosting final accuracy. Our process-aware framework sets a new SOTA of 95.48% on MQUAKE-CF-3k (+5.28%), demonstrating that for complex tasks, aligning the reasoning process is essential for building trustworthy LLMs.

Purpose: The security of autonomous vehicle networks is facing major challenges, owing to the complexity of sensor integration, real-time performance demands, and distributed communication protocols that expose vast attack surfaces around both individual and network-wide safety. Existing security schemes are unable to provide sub-10 ms (milliseconds) anomaly detection and distributed coordination of large-scale networks of vehicles within an acceptable safety/privacy framework. Method: This paper introduces a three-tier hybrid security architecture HA VEN (Hierarchical Autonomous Vehicle Enhanced Network), which decouples real-time local threat detection and distributed coordination operations. It incorporates a light ensemble anomaly detection model on the edge (first layer), Byzantine-fault-tolerant federated learning to aggregate threat intelligence at a regional scale (middle layer), and selected blockchain mechanisms (top layer) to ensure critical security coordination. Result: Extensive experimentation is done on a real-world autonomous driving dataset. Large-scale simulations with the number of vehicles ranging between 100 and 1000 and different attack types, such as sensor spoofing, jamming, and adversarial model poisoning, are conducted to test the scalability and resiliency of HA VEN. Conclusion: The proposed framework overcomes the important tradeoff between real-time safety obligation and distributed security coordination with novel three-tiered processing. The scalable architecture of HAVEN is shown to provide great improvement in detection accuracy as well as network resilience over other methods. Introduction The unprecedented rise of autonomous vehicles (A V) has altered the transport industry by providing unparalleled connectivity, intelligence, and an accessible transportation medium to people. These vehicles process mul-timodal sensor data collected from LiDAR, cameras, radar, and GPS/IMU sensors networked on a CAN (Controller Area Network) bus, generating terabytes of data in a day that require real-time analysis for safe operation [1, 2, 3]. However, the automotive systems are marred by state-of-the-art security threats characterized by safety-critical domains. Further, the distributed nature of vehicular networks poses significantly higher computational complexity and inherent difficulties while developing efficient cyberse-curity frameworks for detecting wide-array of threats in real-time.

This article revisits the 20-year-old neural fitted Q-iteration (NFQ) algorithm on its classical CartPole benchmark. NFQ was a pioneering approach towards modern Deep Reinforcement Learning (Deep RL) in applying multi-layer neural networks to reinforcement learning for real-world control problems. We explore the algorithm's conceptual simplicity and its transition from online to batch learning, which contributed to its stability. Despite its initial success, NFQ required extensive tuning and was not easily reproducible on real-world control problems. We propose a modernized variant NFQ2.0 and apply it to the CartPole task, concentrating on a real-world system build from standard industrial components, to investigate and improve the learning process's repeatability and robustness. Through ablation studies, we highlight key design decisions and hyperparameters that enhance performance and stability of NFQ2.0 over the original variant. Finally, we demonstrate how our findings can assist practitioners in reproducing and improving results and applying deep reinforcement learning more effectively in industrial contexts.

Web Application Firewalls are crucial for protecting web applications against a wide range of cyber threats. Traditional Web Application Firewalls often struggle to effectively distinguish between malicious and legitimate traffic, leading to limited efficacy in threat detection. To overcome these limitations, this paper proposes an Adaptive Dual-Layer WAF employing a two-layered Machine Learning model designed to enhance the accuracy of anomaly and threat detection. The first layer employs a Decision Tree (DT) algorithm to detect anomalies by identifying traffic deviations from established normal patterns. The second layer employs Support Vector Machine to classify these anomalies as either threat anomalies or benign anomalies. Our Adaptive Dual Layer WAF incorporates comprehensive data pre-processing and feature engineering techniques and has been thoroughly evaluated using five large benchmark datasets. Evaluation using these datasets shows that ADL WAF achieves a detection accuracy of 99.88% and a precision of 100%, significantly enhancing anomaly detection and reducing false positives. These findings suggest that integrating machine learning techniques into WAFs can substantially improve web application security by providing more accurate and efficient threat detection.

Context: Large language models (LLMs) are released faster than users' ability to evaluate them rigorously. When LLMs underpin research, such as identifying relevant literature for systematic reviews (SRs), robust empirical assessment is essential. Objective: We identify and discuss key challenges in assessing LLM performance for selecting relevant literature, identify good (evaluation) practices, and propose recommendations. Method: Using a recent large-scale study as an example, we identify problems with the use of traditional metrics for assessing the performance of Gen-AI tools for identifying relevant literature in SRs. We analyzed 27 additional papers investigating this issue, extracted the performance metrics, and found both good practices and widespread problems, especially with the use and reporting of performance metrics for SR screening. Results: Major weaknesses included: i) a failure to use metrics that are robust to imbalanced data and do not directly indicate whether results are better than chance, e.g., the use of Accuracy, ii) a failure to consider the impact of lost evidence when making claims concerning workload savings, and iii) pervasive failure to report the full confusion matrix (or performance metrics from which it can be reconstructed) which is essential for future meta-analyses. On the positive side, we extract good (evaluation) practices on which our recommendations for researchers and practitioners, as well as policymakers, are built. Conclusions: SR screening evaluations should prioritize lost evidence/recall alongside chance-anchored and cost-sensitive Weighted MCC (WMCC) metric, report complete confusion matrices, treat unclassifiable outputs as referred-back positives for assessment, adopt leakage-aware designs with non-LLM baselines and open artifacts, and ground conclusions in cost-benefit analysis where FNs carry higher penalties than FPs.