arXiv subjects

Gurpreet Singh

Publications and source records attributed to Gurpreet Singh.

At least 19 recordsLinked to original sources

arXiv2026

A Comprehensive Dataset for Human vs. AI Generated Text Detection

The rapid advancement of large language models (LLMs) has led to increasingly human-like AI-generated text, raising concerns about content authenticity, misinformation, and trustworthiness. Addressing the challenge of reliably detecting AI-generated text and attributing it to specific models requires large-scale, diverse, and well-annotated datasets. In this work, we present a comprehensive dataset comprising over 73,193 text samples that combine authentic New York Times articles with synthetic versions generated by multiple state-of-the-art LLMs including Gemma-2-9b, Mistral-7B, Qwen-2-72B, LLaMA-8B, Yi-Large, and GPT-4-o. The dataset provides original article abstracts as prompts, full human-authored narratives. We establish baseline results for two key tasks: distinguishing human-written from AI-generated text, achieving an accuracy of 58.35\%, and attributing AI texts to their generating models with an accuracy of 8.92\%. By bridging real-world journalistic content with modern generative models, the dataset aims to catalyze the development of robust detection and attribution methods, fostering trust and transparency in the era of generative AI. Our dataset is available at: https://huggingface.co/datasets/Rajarshi-Roy-research/Defactify_Text_Dataset

cs.CL↗

arXiv2026

A Comprehensive Dataset for Human vs. AI Generated Image Detection

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.

cs.CV↗

arXiv2026

Findings of the Counter Turing Test: AI-Generated Text Detection

The growing capability of large language models to produce fluent, contextually coherent text has created mounting pressure on the systems and institutions responsible for ensuring the authenticity 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.

cs.CL↗

arXiv2026

Findings of the Counter Turing Test: AI-Generated Image Detection

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 support both tasks, we employed the MS COCOAI dataset, a benchmark of 96000 real and synthetic images generated by five state-of-the-art models alongside real images from MS COCO. 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.

cs.CV↗

arXiv2026

Quiescent and flaring states of three active stars: V834 Tau, LQ Hya, and BY Dra

We present a detailed X-ray study of the quiescent and flaring coronae of three active main-sequence K-type stars, V834 Tau, LQ Hya, and BY Dra, using \textit{XMM-Newton} observations. The quiescent coronae are well described by two-temperature thermal plasma models, with cool and hot components at 0.26-0.30 keV and 0.93-1.01 keV, respectively. Despite similar coronal temperatures, X-ray luminosities (10$^{29.18\mbox{--}29.75}$ erg s$^{-1}$) and overall abundances, the relative emission measures of the cool and hot components differ among the stars. High-resolution spectroscopy reveals significant iron depletion by factors of 5-10 relative to photospheric values, and an inverse first-ionisation-potential effect in all three stellar coronae. Six energetic flares are detected, with peak temperatures of 30 -- 133 MK and released energies of $0.6\mbox{--}4.2\times$10$^ {33}$ erg, classifying them as superflares. Most flares exhibit decay times roughly twice their rise times, although one event shows a decay phase nearly twenty times longer than its rise. Time-resolved spectroscopy and loop scaling laws yield flare parameters consistent with previous studies of active stars. LQ Hya displays recurrent superflares at the same rotational phase across observations separated by six months, suggesting a long-lived, complex magnetic field structure. These results provide insights into the magnetic activity and flare energetics in active stars, and their implications for stellar and exoplanetary environments.

astro-ph.SR↗

arXiv2026

Graphene-based technologies for energy applications, challenges and perspectives

Here we report on technology developments implemented into the Graphene Flagship European project for the integration of graphene and graphene-related materials (GRMs) into energy application devices. Many of the technologies investigated so far aim at producing composite materials associating graphene or GRMs with either metal or semiconducting nanocrystals or other carbon nanostructures (e.g., CNT, graphite). These composites can be used favourably as hydrogen storage materials or solar cell absorbers. They can also provide better performing electrodes for fuel cells, batteries, or supercapacitors. For photovoltaic (PV) electrodes, where thin layers and interface engineering are required, surface technologies are preferred. We are using conventional vacuum processes to integrate graphene as well as radically new approaches based on laser irradiation strategies. For each application, the potential of implemented technologies is then presented on the basis of selected experimental and modelling results. It is shown in particular how some of these technologies can maximize the benefit taken from GRM integration. The technical challenges still to be addressed are highlighted and perspectives derived from the running works emphasized.

cond-mat.mtrl-sci↗

arXiv2025

Optical microcavity with a pair of suspended resonant mirrors

We report on the realization of an optical microcavity consisting in the plane-plane arrangement of two suspended resonant mirrors possessing spectrally overlapping high-quality factor internal resonances. We first investigate its generic transmission spectra as the cavity length is varied on the basis of a simple linear Fabry-Perot model, compare them with those of broadband mirror cavities or Fano cavities possessing a single resoannt mirror, and then present an experimental realization using a pair of highly pretensioned, ultrathin silicon nitride films patterned with one-dimensional photonic crystal structures.

physics.optics↗

arXiv2025

Sentinel: Dynamic Knowledge Distillation for Personalized Federated Intrusion Detection in Heterogeneous IoT Networks

Federated learning (FL) offers a privacy-preserving paradigm for machine learning, but its application in intrusion detection systems (IDS) within IoT networks is challenged by severe class imbalance, non-IID data, and high communication overhead.These challenges severely degrade the performance of conventional FL methods in real-world network traffic classification. To overcome these limitations, we propose Sentinel, a personalized federated IDS (pFed-IDS) framework that incorporates a dual-model architecture on each client, consisting of a personalized teacher and a lightweight shared student model. This design effectively balances deep local adaptation with efficient global model consensus while preserving client privacy by transmitting only the compact student model, thus reducing communication costs. Sentinel integrates three key mechanisms to ensure robust performance: bidirectional knowledge distillation with adaptive temperature scaling, multi-faceted feature alignment, and class-balanced loss functions. Furthermore, the server employs normalized gradient aggregation with equal client weighting to enhance fairness and mitigate client drift. Extensive experiments on the IoTID20 and 5GNIDD benchmark datasets demonstrate that Sentinel significantly outperforms state-of-the-art federated methods, establishing a new performance benchmark, especially under extreme data heterogeneity, while maintaining communication efficiency.

cs.LG↗

arXiv2025

RADIANT: Retrieval AugmenteD entIty-context AligNmenT -- Introducing RAG-ability and Entity-Context Divergence

As Large Language Models (LLMs) continue to advance, Retrieval-Augmented Generation (RAG) has emerged as a vital technique to enhance factual accuracy by integrating external knowledge into the generation process. However, LLMs often fail to faithfully integrate retrieved evidence into their generated responses, leading to factual inconsistencies. To quantify this gap, we introduce Entity-Context Divergence (ECD), a metric that measures the extent to which retrieved information is accurately reflected in model outputs. We systematically evaluate contemporary LLMs on their ability to preserve factual consistency in retrieval-augmented settings, a capability we define as RAG-ability. Our empirical analysis reveals that RAG-ability remains low across most LLMs, highlighting significant challenges in entity retention and context fidelity. This paper introduces Radiant (Retrieval AugmenteD entIty-context AligNmenT), a novel framework that merges RAG with alignment designed to optimize the interplay between retrieved evidence and generated content. Radiant extends Direct Preference Optimization (DPO) to teach LLMs how to integrate provided additional information into subsequent generations. As a behavior correction mechanism, Radiant boosts RAG performance across varied retrieval scenarios, such as noisy web contexts, knowledge conflicts, and hallucination reduction. This enables more reliable, contextually grounded, and factually coherent content generation.

cs.CL↗

arXiv2025

DETONATE: A Benchmark for Text-to-Image Alignment and Kernelized Direct Preference Optimization

Alignment is crucial for text-to-image (T2I) models to ensure that generated images faithfully capture user intent while maintaining safety and fairness. Direct Preference Optimization (DPO), prominent in large language models (LLMs), is extending its influence to T2I systems. This paper introduces DPO-Kernels for T2I models, a novel extension enhancing alignment across three dimensions: (i) Hybrid Loss, integrating embedding-based objectives with traditional probability-based loss for improved optimization; (ii) Kernelized Representations, employing Radial Basis Function (RBF), Polynomial, and Wavelet kernels for richer feature transformations and better separation between safe and unsafe inputs; and (iii) Divergence Selection, expanding beyond DPO's default Kullback-Leibler (KL) regularizer by incorporating Wasserstein and R'enyi divergences for enhanced stability and robustness. We introduce DETONATE, the first large-scale benchmark of its kind, comprising approximately 100K curated image pairs categorized as chosen and rejected. DETONATE encapsulates three axes of social bias and discrimination: Race, Gender, and Disability. Prompts are sourced from hate speech datasets, with images generated by leading T2I models including Stable Diffusion 3.5 Large, Stable Diffusion XL, and Midjourney. Additionally, we propose the Alignment Quality Index (AQI), a novel geometric measure quantifying latent-space separability of safe/unsafe image activations, revealing hidden vulnerabilities. Empirically, we demonstrate that DPO-Kernels maintain strong generalization bounds via Heavy-Tailed Self-Regularization (HT-SR). DETONATE and complete code are publicly released.

cs.CV↗

arXiv2025

Alignment Quality Index (AQI) : Beyond Refusals: AQI as an Intrinsic Alignment Diagnostic via Latent Geometry, Cluster Divergence, and Layer wise Pooled Representations

Alignment is no longer a luxury, it is a necessity. As large language models (LLMs) enter high-stakes domains like education, healthcare, governance, and law, their behavior must reliably reflect human-aligned values and safety constraints. Yet current evaluations rely heavily on behavioral proxies such as refusal rates, G-Eval scores, and toxicity classifiers, all of which have critical blind spots. Aligned models are often vulnerable to jailbreaking, stochasticity of generation, and alignment faking. To address this issue, we introduce the Alignment Quality Index (AQI). This novel geometric and prompt-invariant metric empirically assesses LLM alignment by analyzing the separation of safe and unsafe activations in latent space. By combining measures such as the Davies-Bouldin Score (DBS), Dunn Index (DI), Xie-Beni Index (XBI), and Calinski-Harabasz Index (CHI) across various formulations, AQI captures clustering quality to detect hidden misalignments and jailbreak risks, even when outputs appear compliant. AQI also serves as an early warning signal for alignment faking, offering a robust, decoding invariant tool for behavior agnostic safety auditing. Additionally, we propose the LITMUS dataset to facilitate robust evaluation under these challenging conditions. Empirical tests on LITMUS across different models trained under DPO, GRPO, and RLHF conditions demonstrate AQI's correlation with external judges and ability to reveal vulnerabilities missed by refusal metrics. We make our implementation publicly available to foster future research in this area.

cs.CL↗

arXiv2025

Co-optimize condenser water temperature and cooling tower fan using high-fidelity synthetic data

This paper introduces a novel method for optimizing HVAC systems in buildings by integrating a high-fidelity physics-based simulation model with machine learning and measured data. The method enables a real-time building advisory system that provides optimized settings for condenser water loop operation, assisting building operators in decision-making. The building and its HVAC system are first modeled using eQuest. Synthetic data are then generated by running the simulation multiple times. The data are then processed, cleaned, and used to train the machine learning model. The machine learning model enables real-time optimization of the condenser water loop using particle swarm optimization. The results deliver both a real-time online optimizer and an offline operation look-up table, providing optimized condenser water temperature settings and the optimal number of cooling tower fans at a given cooling load. Potential savings are calculated by comparing measured data from two summer months with the energy costs the building would have experienced under optimized settings. Adaptive model refinement is applied to further improve accuracy and effectiveness by utilizing available measured data. The method bridges the gap between simulation and real-time control. It has the potential to be applied to other building systems, including the chilled water loop, heating systems, ventilation systems, and other related processes. Combining physics models, data models, and measured data also enables performance analysis, tracking, and retrofit recommendations.

eess.SY↗

arXiv2025

YINYANG-ALIGN: Benchmarking Contradictory Objectives and Proposing Multi-Objective Optimization based DPO for Text-to-Image Alignment

Precise alignment in Text-to-Image (T2I) systems is crucial to ensure that generated visuals not only accurately encapsulate user intents but also conform to stringent ethical and aesthetic benchmarks. Incidents like the Google Gemini fiasco, where misaligned outputs triggered significant public backlash, underscore the critical need for robust alignment mechanisms. In contrast, Large Language Models (LLMs) have achieved notable success in alignment. Building on these advancements, researchers are eager to apply similar alignment techniques, such as Direct Preference Optimization (DPO), to T2I systems to enhance image generation fidelity and reliability. We present YinYangAlign, an advanced benchmarking framework that systematically quantifies the alignment fidelity of T2I systems, addressing six fundamental and inherently contradictory design objectives. Each pair represents fundamental tensions in image generation, such as balancing adherence to user prompts with creative modifications or maintaining diversity alongside visual coherence. YinYangAlign includes detailed axiom datasets featuring human prompts, aligned (chosen) responses, misaligned (rejected) AI-generated outputs, and explanations of the underlying contradictions.

cs.AI↗

arXiv2025

DPO Kernels: A Semantically-Aware, Kernel-Enhanced, and Divergence-Rich Paradigm for Direct Preference Optimization

The rapid rise of large language models (LLMs) has unlocked many applications but also underscores the challenge of aligning them with diverse values and preferences. Direct Preference Optimization (DPO) is central to alignment but constrained by fixed divergences and limited feature transformations. We propose DPO-Kernels, which integrates kernel methods to address these issues through four key contributions: (i) Kernelized Representations with polynomial, RBF, Mahalanobis, and spectral kernels for richer transformations, plus a hybrid loss combining embedding-based and probability-based objectives; (ii) Divergence Alternatives (Jensen-Shannon, Hellinger, Renyi, Bhattacharyya, Wasserstein, and f-divergences) for greater stability; (iii) Data-Driven Selection metrics that automatically choose the best kernel-divergence pair; and (iv) a Hierarchical Mixture of Kernels for both local precision and global modeling. Evaluations on 12 datasets demonstrate state-of-the-art performance in factuality, safety, reasoning, and instruction following. Grounded in Heavy-Tailed Self-Regularization, DPO-Kernels maintains robust generalization for LLMs, offering a comprehensive resource for further alignment research.

cs.LG↗

arXiv2024

Iterative NLP Query Refinement for Enhancing Domain-Specific Information Retrieval: A Case Study in Career Services

Retrieving semantically relevant documents in niche domains poses significant challenges for traditional TF-IDF-based systems, often resulting in low similarity scores and suboptimal retrieval performance. This paper addresses these challenges by introducing an iterative and semi-automated query refinement methodology tailored to Humber College's career services webpages. Initially, generic queries related to interview preparation yield low top-document similarities (approximately 0.2--0.3). To enhance retrieval effectiveness, we implement a two-fold approach: first, domain-aware query refinement by incorporating specialized terms such as resources-online-learning, student-online-services, and career-advising; second, the integration of structured educational descriptors like "online resume and interview improvement tools." Additionally, we automate the extraction of domain-specific keywords from top-ranked documents to suggest relevant terms for query expansion. Through experiments conducted on five baseline queries, our semi-automated iterative refinement process elevates the average top similarity score from approximately 0.18 to 0.42, marking a substantial improvement in retrieval performance. The implementation details, including reproducible code and experimental setups, are made available in our GitHub repositories \url{https://github.com/Elipei88/HumberChatbotBackend} and \url{https://github.com/Nisarg851/HumberChatbot}. We also discuss the limitations of our approach and propose future directions, including the integration of advanced neural retrieval models.

cs.IR↗

arXiv2024

Existence results for singular p-biharmonic problem with Hardy potential and critical Hardy-Sobolev exponent

In this article, we consider the singular $p-$biharmonic problem involving Hardy potential and citical Hardy-Sobolev exponent. We study the existence of ground state solutions and least energy sign-changing solutions of the following problem \begin{equation*} Δ_{p}^{2} u -λ_{1} \frac{|u|^{p-2}u}{|x|^{2p}}= \frac{|u|^{p_{*}(α)-2}}{|x|^α}u+λ_{2}\Big(|x|^{-β}*|u|^{q}\Big)|u|^{q-2}u \quad\mbox{ in }\R^{N}, \end{equation*} where $p>2$, $2 0$, $λ_{2} \in \R$, $α, β\in (0,N)$, $p_{*}(α)=\frac{p(N-α)}{N-2p}$ and $N\geq 5$. Firstly, we study existence of ground state solutions by using the minimization method on the associated Nehari manifold. Then, we investigate the least energy sign-changing solutions by considering the Nehari nodal set.

math.AP↗

arXiv2024

Coronal modulation in the ultra-fast rotator LO Peg

We present coronal imaging of the ultra-fast rotator, LO Peg, using the X-ray observations from XMM-Newton. The X-ray light curves show one strong flare at the end of observation, as reported in an earlier study. On removal of flaring events, the quiescent state light curve shows rotational modulations, which are modelled using a maximum likelihood model. The results obtained from modelling show the corona of LO Peg is not uniform. Active regions are concentrated around two longitudes, where one active region appears to be compact. The large coronal area that covers almost 60 degrees along longitude from the poles to the equator does not consist of active regions.

astro-ph.SR↗

arXiv2024

Highly reflective and high-$Q$ thin resonant subwavelength gratings

We theoretically investigate the design of thin subwavelength gratings possessing high-reflectivity and high-$Q$ resonances when illuminated at normal incidence by a Gaussian beam. We compare the performances of single-period and dual-period rectangular gratings using Finite Element Method-based optimization and predict one to two orders of magnitude improvement in their transmission loss-linewidth product, which is the relevant figure of merit for e.g. resonant mirror-based microcavity applications.

physics.optics↗