Large Language Models (LLMs) have made significant progress in reasoning tasks across various domains such as mathematics and coding. However, their performance deteriorates in tasks requiring rich socio-cultural knowledge and diverse local contexts, particularly those involving Indian Culture. Existing Cultural benchmarks are (i) Manually crafted, (ii) contain single-hop questions testing factual recall, and (iii) prohibitively costly to scale, leaving this deficiency largely unmeasured. To address this, we introduce VIRAASAT, a novel, semi-automated multi-hop approach for generating cultural specific multi-hop Question-Answering dataset for Indian culture. VIRAASAT leverages a Knowledge Graph comprising more than 700 expert-curated cultural artifacts, covering 13 key attributes of Indian culture (history, festivals, etc). VIRAASAT spans all 28 states and 8 Union Territories, yielding more than 3,200 multi-hop questions that necessitate chained cultural reasoning. We evaluate current State-of-the-Art (SOTA) LLMs on VIRAASAT and identify key limitations in reasoning wherein fine-tuning on Chain-of-Thought(CoT) traces fails to ground and synthesize low-probability facts. To bridge this gap, we propose a novel framework named Symbolic Chain-of-Manipulation (SCoM). Adapting the Chain-of-Manipulation paradigm, we train the model to simulate atomic Knowledge Graph manipulations internally. SCoM teaches the model to reliably traverse the topological structure of the graph. Experiments on Supervised Fine-Tuning (SFT) demonstrate that SCoM outperforms standard CoT baselines by up to 20

Multimodal generative AI systems like Stable Diffusion, DALL-E, and MidJourney have fundamentally changed how synthetic images are created. These tools drive innovation but also enable the spread of misleading content, false information, and manipulated media. As generated images become harder to distinguish from photographs, detecting them has become an urgent priority. To combat this challenge, we release MS COCOAI, a novel dataset for AI generated image detection consisting of 96000 real and synthetic datapoints, built using the MS COCO dataset. To generate synthetic images, we use five generators: Stable Diffusion 3, Stable Diffusion 2.1, SDXL, DALL-E 3, and MidJourney v6. Based on the dataset, we propose two tasks: (1) classifying images as real or generated, and (2) identifying which model produced a given synthetic image. The dataset is available at https://huggingface.co/datasets/Rajarshi-Roy-research/Defactify_Image_Dataset.

Background Mild cognitive impairment (MCI) is a prevalent condition among older adults and a potential marker for dementia. The current challenge lies in diagnosing MCI among healthy older populations. This diagnosis typically requires extensive labor-intensive neuropsychological evaluation using tools like the Mini-Mental State Examination (MMSE) or the MoCA (Montreal Cognitive Assessment) based on specific diagnostic criteria. Objective This study used knowledge-guided machine learning (ML) algorithms and large language models (LLMs) to build diagnostic models. Our approach generates a clinician-guided classification by augmenting LLMs with external knowledge to predict levels of MCI by using the verbal text from picture description tasks. Methods The models used language features from two picture description tasks, along with demographic features. They aimed to distinguish between three levels of MCI (MCI, possible MCI, and healthy group). The dataset exhibits class imbalance, with relatively fewer MCI and possible-MCI cases compared to healthy participants. This imbalance is addressed using SMOTE and stratified cross-validation. We utilized the cognitive cross-domain attention model (CCDA) to integrate the attention mechanism of diverse types of information into our training process, improving performance. Results We demonstrate the efficacy of machine learning, large language models (LLMs), and knowledge-integrated LLMs built on semantic, syntactic, lexical, fluency, and demographic features to identify different levels of cognitive decline from the analysis of verbal utterances. Our CCDA model detected MCI from the participant input, aided by an external attention mechanism. Statistical analysis was conducted using a significance threshold of α = 0.05. Results that do not meet this threshold are interpreted as indicative trends rather than statistically significant findings. An ablation study (Table 6) showed the impact of the attention mechanism and LLM approach on performance. The proposed CCDA model achieved a mean AUC of 0.81 (95% CI: 0.77–0.85) and an F1 score of 0.73 (95% CI: 0.69–0.76) using stratified 10-fold cross-validation. Conclusion Our knowledge-augmented approach compared favorably to contemporary LLM approaches, indicating the promise of knowledge-augmented learning in detecting MCI. This framework can support early, non-invasive screening for cognitive decline in telehealth and resource-limited settings, assisting clinicians in identifying at-risk individuals for further evaluation.

Autoregressive language models frequently degrade during long-horizon generation, producing repetitive text, losing instruction adherence, and exhibiting unstable entropy. Despite the prevalence of these failures, practitioners lack online diagnostics to detect them in real time as they occur. We formalize this degradation as cognitive fatigue, a measurable generation-time state characterized by decay in attention to the original prompt, representational drift, and entropy miscalibration. We introduce the Fatigue Index (FI), a lightweight, model-agnostic diagnostic that aggregates these three signals under explicit axioms (monotonicity, boundedness, interpretability), enabling reliable runtime monitoring. Across nine models (1B–13B parameters), FI trajectories exhibit structured temporal dynamics, predict task degradation (AUROC = 0.95) and repetition (ρ = 0.94), and reveal non-monotonic scaling behavior: instruction-tuned models below 3B exhibit faster collapse than base models, with this trend reversing at 7B. Stress analyses further show that FI onset accelerates under longer contexts, middle-positioned evidence, and reduced numerical precision. These results establish cognitive fatigue as a coherent and measurable phenomenon, and position FI as a principled tool for runtime reliability monitoring in production LLM systems.

Multiple-choice QA benchmarks usually evaluate small language models (SLMs) as direct answerers, but deployed language-model systems increasingly rely on external scaffolds such as tools, code, and repeated model calls. We introduce Code-Guided Reasoning (CGR), an evaluation protocol and generated-program resource for measuring when executable reasoning scaffolds improve SLM performance on MCQA tasks. CGR standardizes six components: a normalized item interface, a direct solver prompt, a generator prompt, a Python scaffold, solver-call and extraction helpers, and a three-channel result record. On 20,498 retained result rows from a locally prepared MCQA bundle and six metadata-registered solver models, the observed non-zero-baseline partition shows 66.21

The rapid advancements in generative AI technologies, such as Stable Diffusion, DALL-E, and Midjourney, have significantly transformed the creation of synthetic visual content. While these models enable innovation across industries, they also pose serious challenges, including misinformation, disinformation, and biased content generation. The increasing realism of AI-generated images makes their detection a pressing concern for researchers, policymakers, and industry stakeholders. In this paper, we present the findings of the Defactify 4.0 workshop, which introduced the Counter Turing Test (CT2) for AI-Generated Image Detection. The competition consisted of two key tasks: (1) binary classification of images as either AI-generated or real and (2) identification of the specific generative model responsible for an AI-generated image. To facilitate this, we developed the MS COCOAI dataset, consisting of 50,000 synthetic images from multiple generative models alongside real-world images from the MS COCO dataset. Participants employed diverse detection strategies, including convolutional neural networks (CNNs), Vision Transformers (ViTs), frequency-based analysis, contrastive learning, and multimodal techniques. The results demonstrated that while AI-generated images can be detected with high accuracy (F1-score > 0.83), identifying the exact model used remains significantly more challenging (highest F1-score: 0.4986). These findings highlight the need for improved model fingerprinting, adversarial robustness, and real-time detection mechanisms.

Escalating prevalence of mental health issues worldwide has created an unprecedented demand for mental healthcare services, yet the shortage of qualified practitioners limits accessibility for countless individuals in need. AI has emerged as a potential solution to support mental health professionals, offering assistance that goes beyond simple diagnostic aid. This research introduces a novel AI-powered real-time diagnostic support system—MATRIX—for mental healthcare diagnostics designed to interact with users using natural language and utilizing the Patient Health Questionnaire-9 (PHQ-9), a standardized clinical tool for assessing depressive symptoms. The system classifies the interaction into a well-defined checklist and generates the most likely diagnosis through a framework termed X-AI Attribution Reasoning, which provides explainable and attributable diagnostic logic for interdisciplinary clarity. Unlike existing diagnostic support systems that primarily rely on static scoring or predefined rule sets, MATRIX integrates explainable AI (XAI) principles to deliver interpretable reasoning pathways that clinicians can trace and validate. The PHQ-9 implementation within MATRIX has been tested in controlled clinical simulations, confirming its usability and alignment with real-world assessment practices. The system not only accelerates the diagnostic process but also provides transparent explanations, detailed reasoning for such diagnoses, and clinically relevant attributions linked to standard SNOMED Concept IDs, which can be directly utilized by clinicians for documentation, referrals, and electronic health record integration while maintaining data privacy. By offering this level of insight, the system fosters a trustworthy AI-human collaboration that aids clinicians in understanding and validating each diagnostic recommendation. Interpretability within MATRIX is achieved through visual attribution maps and narrative output summaries, ensuring that decision processes remain both transparent and clinically meaningful. The integration of these features enables practitioners to focus on patient care with the assurance that AI-assisted diagnostics align with clinical standards, resulting in reduced time spent per patient and enhanced patient throughput. Our preliminary findings indicate that MATRIX achieves over 89% classification accuracy and high clinician satisfaction in pilot evaluations, demonstrating that AI-driven support systems with explainable, reasonable, and attributable real-time diagnostics can significantly enhance the capacity of mental health services and improve access to timely and effective care for those affected by mental health conditions. This study highlights the essential role of AI in enhancing both the efficiency and trustworthiness of mental health diagnostics in clinical settings, making a compelling case for the integration of AI into modern mental healthcare.

Modern manufacturing systems demand real-time, trustworthy, and interpretable insights into anomalies and their underlying causes. However, conventional pipelines treat anomaly detection, causal inference, and decision-making as siloed tasks, lacking integration, explainability, and adaptability. We present CausalPulse, an intelligent, multi-agent copilot for automated Root Cause Analysis (RCA) in industrial settings. Built on a modular and extensible architecture, the system leverages standard agentic protocols, including Model Context Protocol (MCP), Agent2Agent (A2A), and LangGraph for dynamic tool and agent discovery and seamless orchestration of tasks. Agents dynamically interact to perform data preprocessing, anomaly detection, causal discovery, and root cause analysis through a neurosymbolic workflow that combines symbolic reasoning with neural methods. Intelligent postprocessing pipelines enable automatic chaining of agent tasks, enhancing contextual awareness and adaptability. CausalPulse is evaluated using both an academic public dataset (i.e., Future Factories) and an industrial proprietary dataset (i.e., Planar Oxygen Sensor Element) and shows that the system outperforms traditional baselines in interpretability, trustworthiness, and operational utility.

Modern manufacturing environments demand not only accurate predictions but also interpretable insights to process anomalies, root causes, and potential interventions. Existing AI systems often function as isolated black boxes, lacking the seamless integration of prediction, explanation, and causal reasoning required for a unified decision-support solution. This fragmentation limits their trustworthiness and practical utility in high-stakes industrial environments. In this work, we present CausalTrace, a neurosymbolic causal analysis module integrated into the SmartPilot industrial CoPilot. CausalTrace performs data-driven causal analysis enriched by industrial ontologies and knowledge graphs, including advanced functions such as causal discovery, counterfactual reasoning, and root cause analysis (RCA). It supports real-time operator interaction and is designed to complement existing agents by offering transparent, explainable decision support. We conducted a comprehensive evaluation of CausalTrace using multiple causal assessment methods and the C3AN framework (i.e. Custom, Compact, Composite AI with Neurosymbolic Integration), which spans principles of robustness, intelligence, and trustworthiness. In an academic rocket assembly testbed, CausalTrace achieved substantial agreement with domain experts (ROUGE-1: 0.91 in ontology QA) and strong RCA performance (MAP@3: 94%, PR@2: 97%, MRR: 0.92, Jaccard: 0.92). It also attained 4.59/5 in the C3AN evaluation, demonstrating precision and reliability for live deployment.

LLMs are increasingly being deployed as chatbots, but today's interfaces offer little to no friction: users interact through seamless conversations that conceal when the model is drifting, hallucinating or failing. This lack of transparency fosters blind trust, even as models produce unstable or repetitive outputs. We introduce an interactive demo that surfaces and mitigates cognitive fatigue, a failure mode where LLMs gradually lose coherence during auto-regressive generation. Our system, Chatsparent, instruments real-time, token-level signals of fatigue, including attention-to-prompt decay, embedding drift, and entropy collapse, and visualizes them as a unified fatigue index. When fatigue thresholds are crossed, the interface allows users to activate lightweight interventions such as attention resets, entropy-regularized decoding, and self-reflection checkpoints. The demo streams live text and fatigue signals, allowing users to observe when fatigue arises, how it affects output quality, and how interventions restore stability. By turning passive chatbot interaction into an interactive diagnostic experience, our system empowers users to better understand LLM behavior while improving reliability at inference time.

Modern manufacturing environments demand real-time, trustworthy, and interpretable root-cause insights to sustain productivity and quality. Traditional analytics pipelines often treat anomaly detection, causal inference, and root-cause analysis as isolated stages, limiting scalability and explainability. In this work, we present CausalPulse, an industry-grade multi-agent copilot that automates causal diagnostics in smart manufacturing. It unifies anomaly detection, causal discovery, and reasoning through a neurosymbolic architecture built on standardized agentic protocols. CausalPulse is being deployed in a Robert Bosch manufacturing plant, integrating seamlessly with existing monitoring workflows and supporting real-time operation at production scale. Evaluations on both public (Future Factories) and proprietary (Planar Sensor Element) datasets show high reliability, achieving overall success rates of 98.0% and 98.73%. Per-criterion success rates reached 98.75% for planning and tool use, 97.3% for self-reflection, and 99.2% for collaboration. Runtime experiments report end-to-end latency of 50-60 s per diagnostic workflow with near-linear scalability (R2=0.97), confirming real-time readiness. Comparison with existing industrial copilots highlights distinct advantages in modularity, extensibility, and deployment maturity. These results demonstrate how CausalPulse's modular, human-in-the-loop design enables reliable, interpretable, and production-ready automation for next-generation manufacturing.

As Vision Language Models (VLMs) become integral to real-world applications, understanding their demographic biases is critical. We introduce GRAS, a benchmark for uncovering demographic biases in VLMs across gender, race, age, and skin tone, offering the most diverse coverage to date. We further propose the GRAS Bias Score, an interpretable metric for quantifying bias. We benchmark five state-of-the-art VLMs and reveal concerning bias levels, with the least biased model attaining a GRAS Bias Score of only 2 out of 100. Our findings also reveal a methodological insight: evaluating bias in VLMs with visual question answering (VQA) requires considering multiple formulations of a question. Our code, data, and evaluation results are publicly available.

Multimodal generative models produce fluent outputs but remain unreliable when generation must respect structured, domain-specific, or safety-critical knowledge. Existing methods incorporate knowledge through mechanisms such as prompt augmentation, guidance, latent editing, or fine-tuning, yet they are typically categorized by technique rather than by the component of the generative process they modify. We argue that knowledge infusion in iterative generative models is fundamentally anintervention-layer problem. Since thegenerative process unfolds as a trajectory of internal states, knowledge can act on four structurally distinct components of this process: the input/output boundary, the transition function, the intermediate state, and the model parameters. This maps to four intervention layers: surface, trajectory, latent, and parametric infusion. We instantiate the framework in diffusion models, map representative methods to all four layers, and derive design principles for multi-layer composition. In a controlled safety-alignment experiment using a multimodal knowledge graph with two diffusion backbones, we implement three of the four layers cumulatively, surface (input-side and output-side) and trajectory–latent (mid-generation). We show empirically that each additional layer addresses failure classes that prior layers cannot reach, reducing knowledge-violating outputs by 70.97

The deployment of large language models (LLMs) in specialized domains like medical diagnostics and financial advisory necessitates evaluating capabilities beyond general knowledge. Traditional question-answering benchmarks often fail to capture the nuanced contextual grounding, user awareness, and domain understanding these fields require. To address this, we introduce CANDI-QA (Contextual Alignment for Niche Domains Question Answering), a novel dataset evaluating LLMs on delivering accurate, context-sensitive, and user-aligned answers in specialized settings. CANDI-QA features expert-curated question-answer pairs structured into two categories: (1) Information Assistance Questions, which are direct, factual queries requiring precise extraction, and (2) Applied Inference Questions, which are multi-hop reasoning tasks needing situational inference to generate actionable insights. We evaluate over ten diverse language models, from compact open-source to state-of-the-art proprietary systems. As a robust baseline, we present MTSS-Net, a lightweight neuro-symbolic framework combining neural retrieval with rule-based reasoning. Our findings highlight the profound challenges of achieving contextual alignment in niche domains, revealing the limitations of current LLMs without enhanced contextual or symbolic integration. Ultimately, CANDI-QA serves as a critical benchmark for advancing research in context-aware language models, stimulating the development of robust, trustworthy AI for high-stakes domains.

Several studies have evaluated the ability of Large Language Models (LLMs) for meal planning, yielding positive outcomes. These models can process natural language inputs and leverage learned knowledge from their pretraining to generate meal plans. In this work, we investigate the ability of LLMs to analyze the suitability of given recipes for diabetes. The primary challenge for LLMs is to retrieve relevant dietary guidelines for diabetes, decompose recipes into ingredients and cooking methods, and apply these guidelines to determine the recipe's suitability. To study these challenges, we employ three kinds of prompts namely, (i) Direct Query Prompt (ii) Context-Guided Prompt, and (iii) Exemplary Context Prompt that incorporate different levels of diabetes dietary guidelines from medical sources. We introduce a benchmark dataset curated for this investigation consisting of 7607 recipes that include 3807 recipes suitable for diabetes and 3800 recipes not suitable for diabetes. Our results demonstrate that most LLMs are cautious in predicting recipes as suitable to prevent detrimental outcomes. Further, the models that can reason using the dietary guidelines performed better in predicting the suitability of recipes for diabetes. Overall, Mistral-7B and Llama 70B showed superior performance to their counterparts.

While tool-augmented Large Language Models have significantly improved multi-step reasoning in quantitative STEM tasks, a critical residual failure mode remains: intermediate reasoning steps that are syntactically well-formed, mathematically executable, and unit-consistent, yet contextually ungrounded. Current approaches either rely on formal verifiers that cannot assess semantic intent, or burden Process Reward Models (PRMs) with the dual task of checking both arithmetic and logic. In this paper, we propose a neuro-symbolic framework that cleanly decouples reasoning into two formal dimensions: Symbolic Validity (V) and Semantic Groundedness (G). We guarantee V by construction using a deterministic symbolic verifier acting as a hard filter. To assess G, we train a PRM conditionally on the verifier-accepted manifold. To train this PRM efficiently, we introduce Counterfactual Symbolic Perturbation (CSP), a novel data synthesis strategy that algorithmically generates constraint-preserving hard negatives (steps that perfectly pass the verifier but are logically flawed). At inference, we deploy a verifier-first constrained search that guarantees execution consistency for verifier-covered operations while relying on the PRM solely to rank semantic grounding. By targeting the exact residual error class of strong tool-using LLMs, our method significantly improves reasoning reliability without the sprawling heuristics of prior frameworks.

AI-driven education platforms have made some progress in personalisation, yet most remain constrained to static adaptation—predefined quizzes, uniform pacing, or generic feedback—limiting their ability to respond to learners’ evolving understanding. This shortfall highlights the need for systems that are both context-aware and adaptive in real time. We introduce PAL (Personal Adaptive Learner), an AI-powered platform that transforms lecture videos into interactive learning experiences. PAL continuously analyzes multimodal lecture content and dynamically engages learners through questions of varying difficulty, adjusting to their responses as the lesson unfolds. At the end of a session, PAL generates a personalized summary that reinforces key concepts while tailoring examples to the learner’s interests. By uniting multimodal content analysis with adaptive decision-making, PAL contributes a novel framework for responsive digital learning. Our work demonstrates how AI can move beyond static personalization toward real-time, individualized support, addressing a core challenge in AI-enabled education.

The rapid proliferation of AI-generated text has introduced significant challenges in maintaining the integrity of digital content. Advanced generative models such as GPT-4, Claude 3.5, and Llama can produce highly coherent and human-like text, making it increasingly difficult to differentiate between human-written and AI-generated content. While these models have transformative applications, their misuse has raised concerns about misinformation, biased narratives, and security threats. This paper provides a comprehensive analysis of state-of-the-art AI-generated text detection techniques and evaluates their effectiveness through the Counter Turing Test (CT2) shared tasks. Task A (Binary Classification) required participants to distinguish between human-written and AI-generated text, while Task B (Model Attribution) focused on identifying the specific language model responsible for generating a given text. The results demonstrated high performance in binary classification, with the top system achieving an F1 score of 1.0000, but significantly lower scores in model attribution, where the best system achieved 0.9531, highlighting the increased complexity of this task. The top-performing teams leveraged fine-tuned transformer models, ensemble learning, and hybrid detection approaches, with DeBERTa-based and BART-based methods demonstrating strong results. However, the lower scores in Task B underscore the challenges of distinguishing outputs from different LLMs, necessitating further research into adversarial robustness, feature extraction, and cross-domain generalization.

The innovations of Industry 4.0 have revolutionized the capabilities of modern manufacturing systems by integrating information technology (IT) and operational technology (OT). This convergence has unlocked unprecedented manufacturing potential, enabling advanced intelligence and autonomy. However, a critical challenge remains: the lack of transparency and explainability in the decision-making processes of these systems. Machine learning models, which are at the core of many Industry 4.0 advancements, often function as 'black boxes', providing decisions without clear reasoning behind them. This paper addresses this challenge by introducing a Knowledge Graph-based explainability framework to enhance the reasoning behind machine learning outputs. A process ontology was developed and applied to a robotic assembly line to provide clear reasoning behind the decisions of a classification algorithm. This framework provides technicians explanations behind the prediction of the classifier. This paper details the development of the framework and its successful deployment on a robotic assembly line. DrCIF was chosen as the classifier that identifies defective assembled rockets with an F1 score of 99% after training and testing. The output of this classifier is then explained through the deployed ontology, demonstrating its potential to bridge the gap between advanced manufacturing intelligence and decision-making transparency.

Retrieval-augmented generation (RAG) has made significant strides in overcoming key limitations of large language models, such as hallucination, lack of contextual grounding, and issues with transparency. This new framework aims to answer two primary questions: 1) Can retrievers provide a clear and interpretable basis for document selection? 2) Can symbolic knowledge enhance the clarity of the retrieval process? We propose three methods to improve this integration. The first is modulation augmented retrieval which employs modulation networks to refine query embeddings using interpretable symbolic features, thereby making document matching more explicit. The second is KG-Path RAG, which enhances queries by traversing knowledge graphs to improve overall retrieval quality and interpretability. Finally, process knowledge-infused RAG utilizes domain-specific tools to reorder retrieved content based on validated workflows. Preliminary results from mental health risk assessment tasks indicate that this neurosymbolic approach enhances both transparency and overall performance.