arXiv subjects

Gaurav Kumar Gupta

Publications and source records attributed to Gaurav Kumar Gupta.

At least 19 recordsLinked to original sources

arXiv2026

Benchmarking Local LLMs for Natural-Language-to-SQL Querying in Biopharmaceutical Manufacturing: An Empirical Benchmark on Consumer-Grade Hardware

Biopharmaceutical manufacturing organizations operate under regulatory frameworks such as FDA guidance, EU Good Manufacturing Practice (GMP), and the EU AI Act, which can restrict the use of cloud-based artificial intelligence systems. Locally deployed large language models (LLMs) offer a privacy-preserving alternative, but their suitability for pharmaceutical manufacturing tasks remains underexplored. This study evaluates four open-source LLMs (Qwen 2.5 Coder 7B, Llama 3.1 8B, Mistral 7B, and Meditron 7B) deployed locally via Ollama for natural-language-to-SQL generation over a pharmaceutical manufacturing database. A FastAPI-based evaluation platform, PharmaBatchDB AI, was developed using a synthetic Microsoft SQL Server database containing approximately 63,000 records across Batch, Manufacturing Execution System (MES), and Clean-In-Place (CIP) modules. Models were benchmarked on 60 domain-specific natural-language questions using metrics including SQL extraction rate, SQL compliance, factual consistency, ROUGE-L, hallucination rate, throughput, and latency. Qwen 2.5 Coder 7B, Llama 3.1 8B, and Mistral 7B generated SQL for all evaluation tasks, while Meditron 7B failed on nearly all tasks due to context-window limitations and poor SQL generation capability. Llama 3.1 8B achieved the highest SQL compliance, whereas Qwen 2.5 Coder 7B achieved the strongest overall text similarity and factual consistency. Performance differences between the two leading models were not statistically significant. The results show that code-tuned general-purpose LLMs outperform a domain-specific biomedical model on structured query generation for pharmaceutical manufacturing data. Although fully local, GxP-aligned NLQ systems are feasible on consumer hardware, current performance levels still require human oversight and downstream validation for regulated use.

cs.CL↗

arXiv2026

Graph-Augmented Retrieval for Cross-Entity Financial Sentiment Analysis: A Comparative Study

Retrieval-Augmented Generation (RAG) has become foundational for grounding large language models in domain-specific corpora, yet conventional vector-based RAG systems are fundamentally limited in their ability to capture the structured, multi-entity relationships that underpin financial market analysis. This paper presents a comprehensive comparative study of a novel two-hop Graph-RAG architecture versus a standard vector-only baseline for cross-entity financial sentiment analysis. Our system constructs a sentiment-weighted knowledge graph of 59 equity entities from 255 news articles covering 10 major technology stocks, then augments dense retrieval with intensity-filtered graph traversal over INFLUENCES edges to surface relational evidence inaccessible to vector search alone. We evaluate both architectures on 100 grounded queries (30 Direct, 70 Relational) using semantic similarity, entity recall, RAGAS metrics, latency benchmarks, and ablation studies. Graph-RAG achieves a statistically significant improvement in entity recall (+6.4%, p < 0.001, Wilcoxon signed-rank) and delivers substantially more relevant answers for complex multi-entity queries (+11.7% Answer Relevancy), with gains concentrating in relational question types (+16.1%). Critically, these improvements come at no measurable cost to answer quality (delta = +0.001 semantic similarity, Cohen's d = 0.078), with a modest 22.6% increase in mean latency offset by an 80% reduction in latency variance. An ablation study on the graph traversal intensity threshold reveals an inverted-U relationship with answer quality, identifying tau = 0.5 as optimal over the production default of tau = 0.7. These findings characterize a precision-for-coverage trade-off inherent to graph-augmented retrieval and provide actionable architectural guidance for practitioners building RAG systems for multi-entity financial analysis.

cs.CL↗

arXiv2026

A Formal Security Framework for MCP-Based AI Agents: Threat Taxonomy, Verification Models, and Defense Mechanisms

The Model Context Protocol (MCP), introduced by Anthropic in November 2024 and now governed by the Linux Foundation's Agentic AI Foundation, has rapidly become the de facto standard for connecting large language model (LLM)-based agents to external tools and data sources, with over 97 million monthly SDK downloads and more than 177000 registered tools. However, this explosive adoption has exposed a critical gap: the absence of a unified, formal security framework capable of systematically characterizing, analyzing, and mitigating the diverse threats facing MCP-based agent ecosystems. Existing security research remains fragmented across individual attack papers, isolated benchmarks, and point defense mechanisms. This paper presents MCPSHIELD, a comprehensive formal security framework for MCP-based AI agents. We make four principal contributions: (1) a hierarchical threat taxonomy comprising 7 threat categories and 23 distinct attack vectors organized across four attack surfaces, grounded in the analysis of over 177000 MCP tools; (2) a formal verification model based on labeled transition systems with trust boundary annotations that enables static and runtime analysis of MCP tool interaction chains; (3) a systematic comparative evaluation of 12 existing defense mechanisms, identifying coverage gaps across our threat taxonomy; and (4) a defense in depth reference architecture integrating capability based access control, cryptographic tool attestation, information flow tracking, and runtime policy enforcement. Our analysis reveals that no existing single defense covers more than 34 percent of the identified threat landscape, whereas MCPSHIELD's integrated architecture achieves theoretical coverage of 91 percent. We further identify seven open research challenges that must be addressed to secure the next generation of agentic AI systems.

cs.CR↗

arXiv2025

LLM-Based Support for Diabetes Diagnosis: Opportunities, Scenarios, and Challenges with GPT-5

Diabetes mellitus is a major global health challenge, affecting over half a billion adults worldwide with prevalence projected to rise. Although the American Diabetes Association (ADA) provides clear diagnostic thresholds, early recognition remains difficult due to vague symptoms, borderline laboratory values, gestational complexity, and the demands of long-term monitoring. Advances in large language models (LLMs) offer opportunities to enhance decision support through structured, interpretable, and patient-friendly outputs. This study evaluates GPT-5, the latest generative pre-trained transformer, using a simulation framework built entirely on synthetic cases aligned with ADA Standards of Care 2025 and inspired by public datasets including NHANES, Pima Indians, EyePACS, and MIMIC-IV. Five representative scenarios were tested: symptom recognition, laboratory interpretation, gestational diabetes screening, remote monitoring, and multimodal complication detection. For each, GPT-5 classified cases, generated clinical rationales, produced patient explanations, and output structured JSON summaries. Results showed strong alignment with ADA-defined criteria, suggesting GPT-5 may function as a dual-purpose tool for clinicians and patients, while underscoring the importance of reproducible evaluation frameworks for responsibly assessing LLMs in healthcare.

cs.CL↗

arXiv2025

A survey of agent interoperability protocols: Model Context Protocol (MCP), Agent Communication Protocol (ACP), Agent-to-Agent Protocol (A2A), and Agent Network Protocol (ANP)

Large language model powered autonomous agents demand robust, standardized protocols to integrate tools, share contextual data, and coordinate tasks across heterogeneous systems. Ad-hoc integrations are difficult to scale, secure, and generalize across domains. This survey examines four emerging agent communication protocols: Model Context Protocol (MCP), Agent Communication Protocol (ACP), Agent-to-Agent Protocol (A2A), and Agent Network Protocol (ANP), each addressing interoperability in deployment contexts. MCP provides a JSON-RPC client-server interface for secure tool invocation and typed data exchange. ACP defines a general-purpose communication protocol over RESTful HTTP, supporting MIME-typed multipart messages and synchronous and asynchronous interactions. Its lightweight and runtime-independent design enables scalable agent invocation, while features like session management, message routing, and integration with role-based and decentralized identifiers (DIDs). A2A enables peer-to-peer task delegation using capability-based Agent Cards, supporting secure and scalable collaboration across enterprise agent workflows. ANP supports open network agent discovery and secure collaboration using W3C decentralized identifiers DIDs and JSON-LD graphs. The protocols are compared across multiple dimensions, including interaction modes, discovery mechanisms, communication patterns, and security models. Based on the comparative analysis, a phased adoption roadmap is proposed: beginning with MCP for tool access, followed by ACP for structured, multimodal messaging session-aware interaction and both online and offline agent discovery across scalable, HTTP-based deployments A2A for collaborative task execution, and extending to ANP for decentralized agent marketplaces. This work provides a comprehensive foundation for designing secure, interoperable, and scalable ecosystems of LLM-powered agents.

cs.AI↗

arXiv2025

LLMs in Disease Diagnosis: A Comparative Study of DeepSeek-R1 and O3 Mini Across Chronic Health Conditions

Large Language Models (LLMs) are revolutionizing medical diagnostics by enhancing both disease classification and clinical decision-making. In this study, we evaluate the performance of two LLM- based diagnostic tools, DeepSeek R1 and O3 Mini, using a structured dataset of symptoms and diagnoses. We assessed their predictive accuracy at both the disease and category levels, as well as the reliability of their confidence scores. DeepSeek R1 achieved a disease-level accuracy of 76% and an overall accuracy of 82%, outperforming O3 Mini, which attained 72% and 75% respectively. Notably, DeepSeek R1 demonstrated exceptional performance in Mental Health, Neurological Disorders, and Oncology, where it reached 100% accuracy, while O3 Mini excelled in Autoimmune Disease classification with 100% accuracy. Both models, however, struggled with Respiratory Disease classification, recording accuracies of only 40% for DeepSeek R1 and 20% for O3 Mini. Additionally, the analysis of confidence scores revealed that DeepSeek R1 provided high-confidence predictions in 92% of cases, compared to 68% for O3 Mini. Ethical considerations regarding bias, model interpretability, and data privacy are also discussed to ensure the responsible integration of LLMs into clinical practice. Overall, our findings offer valuable insights into the strengths and limitations of LLM-based diagnostic systems and provide a roadmap for future enhancements in AI-driven healthcare.

cs.CL↗

arXiv2025

Diabatic error and propagation of Majorana zero modes in interacting quantum dots systems

Motivated by recent experimental progress in realizing Majorana zero modes (MZMs) using quantum dot systems, we investigate the diabatic errors associated with the movement of those MZMs. The movement is achieved by tuning time-dependent gate potentials applied to individual quantum dots, effectively creating a moving potential wall. To probe the optimized movement of MZMs, we calculate the experimentally accessible local density-of-states and time-dependent fidelity using many-body time-dependent numerical methods. Our analysis reveals that an optimal potential wall height is crucial to preserve the well-localized nature of the MZM during its movement. Moreover, for the first time, we analyze diabatic errors in realistic quantum-dot systems, incorporating the effects of repulsive Coulomb interactions and disorder in both hopping and pairing terms. Additionally, we provide a comparative study of diabatic errors arising from the simultaneous versus sequential tuning of multiple gates during the MZMs movement. Finally, we estimate the time scale required for MZM transfer in a six-quantum-dot system, demonstrating that MZM movement is feasible and can be completed well within the qubit's operational lifetime in practical quantum-dot setups.

cond-mat.mes-hall↗

arXiv2025

Exploring Secure Machine Learning Through Payload Injection and FGSM Attacks on ResNet-50

This paper investigates the resilience of a ResNet-50 image classification model under two prominent security threats: Fast Gradient Sign Method (FGSM) adversarial attacks and malicious payload injection. Initially, the model attains a 53.33% accuracy on clean images. When subjected to FGSM perturbations, its overall accuracy remains unchanged; however, the model's confidence in incorrect predictions notably increases. Concurrently, a payload injection scheme is successfully executed in 93.33% of the tested samples, revealing how stealthy attacks can manipulate model predictions without degrading visual quality. These findings underscore the vulnerability of even high-performing neural networks and highlight the urgency of developing more robust defense mechanisms for security-critical applications.

cs.CR↗

arXiv2024

Phase Diagram of Kane-Mele Hubbard model at small doping

Recent experiment on MoTe$_2$-WTe$_2$ twisted bi-layer demonstrated physics of Kane-Mele (KM) as well as Haldane models. Although topological properties of KM model has been studied extensively, effects of interaction are still less explored beyond half filling. In this work we study the effect of Hubbard interaction in KM model at small hole doping around half filling. We use density matrix renormalization group method to characterize different phase of cylinders with width of $L_y=3$ unit cells. We identify {a possible} superconducting (SC) phase and transition at different strength of spin-orbit coupling (SOC) and show the strong dependence of critical interaction for SC transition. We also calculate the single-particle Green's function, spin-spin and density-density correlations and compare with SC correlation to establish the dominance of SC correlations. Our result highlights striking behaviour of SC transition and its dependence on SOC which could be useful in understanding different ways to achieve unconventional SC.

cond-mat.str-el↗

arXiv2024

The Interplay Between Physical Activity, Protein Consumption, and Sleep Quality in Muscle Protein Synthesis

This systematic review examines the synergistic and individual influences of resistance exercise, dietary protein supplementation, and sleep/recovery on muscle protein synthesis (MPS). Electronic databases such as Scopus, Google Scholar, and Web of Science were extensively used. Studies were selected based on relevance to the criteria and were ensured to be directly applicable to the objectives. Research indicates that a protein dose of 20 to 25 grams maximally stimulates MPS post-resistance training. It is observed that physically frail individuals aged 76 to 92 and middle-aged adults aged 62 to 74 have lower mixed muscle protein synthetic rates than individuals aged 20 to 32. High-whey protein and leucine-enriched supplements enhance MPS more efficiently than standard dairy products in older adults engaged in resistance programs. Similarly, protein intake before sleep boosts overnight MPS rates, which helps prevent muscle loss associated with sleep debt, exercise-induced damage, and muscle-wasting conditions like sarcopenia and cachexia. Resistance exercise is a functional intervention to achieve muscular adaptation and improve function. Future research should focus on variables such as fluctuating fitness levels, age groups, genetics, and lifestyle factors to generate more accurate and beneficial results.

q-bio.TO↗

arXiv2024

Exploring Prompt Engineering: A Systematic Review with SWOT Analysis

In this paper, we conduct a comprehensive SWOT analysis of prompt engineering techniques within the realm of Large Language Models (LLMs). Emphasizing linguistic principles, we examine various techniques to identify their strengths, weaknesses, opportunities, and threats. Our findings provide insights into enhancing AI interactions and improving language model comprehension of human prompts. The analysis covers techniques including template-based approaches and fine-tuning, addressing the problems and challenges associated with each. The conclusion offers future research directions aimed at advancing the effectiveness of prompt engineering in optimizing human-machine communication.

cs.CL↗

arXiv2024

Encouraging Responsible Use of Generative AI in Education: A Reward-Based Learning Approach

This research introduces an innovative mathematical learning approach that integrates generative AI to cultivate a structured learning rather than quick solution. Our method combines chatbot capabilities and generative AI to offer interactive problem-solving exercises, enhancing learning through a stepby-step approach for varied problems, advocating for the responsible use of AI in education. Our approach emphasizes that immediate answers from ChatGPT can impede real learning. We introduce a reward-based system that requires students to solve mathematical problems effectively to receive the final answer. This encourages a progressive learning path from basic to complex problems, rewarding mastery with final solutions. The goal is to transition students from seeking quick fixes to engaging actively in a comprehensive learning experience.

cs.CY↗

arXiv2024

Digital Diagnostics: The Potential Of Large Language Models In Recognizing Symptoms Of Common Illnesses

The recent swift development of LLMs like GPT-4, Gemini, and GPT-3.5 offers a transformative opportunity in medicine and healthcare, especially in digital diagnostics. This study evaluates each model diagnostic abilities by interpreting a user symptoms and determining diagnoses that fit well with common illnesses, and it demonstrates how each of these models could significantly increase diagnostic accuracy and efficiency. Through a series of diagnostic prompts based on symptoms from medical databases, GPT-4 demonstrates higher diagnostic accuracy from its deep and complete history of training on medical data. Meanwhile, Gemini performs with high precision as a critical tool in disease triage, demonstrating its potential to be a reliable model when physicians are trying to make high-risk diagnoses. GPT-3.5, though slightly less advanced, is a good tool for medical diagnostics. This study highlights the need to study LLMs for healthcare and clinical practices with more care and attention, ensuring that any system utilizing LLMs promotes patient privacy and complies with health information privacy laws such as HIPAA compliance, as well as the social consequences that affect the varied individuals in complex healthcare contexts. This study marks the start of a larger future effort to study the various ways in which assigning ethical concerns to LLMs task of learning from human biases could unearth new ways to apply AI in complex medical settings.

cs.CL↗

arXiv2020

Electronic Floquet Liquid Crystals

"Floquet engineering" - designing band structures "on-demand" through the application of coherent time-periodic drives - has recently emerged as a powerful tool for creating new topological and anomalous phases of matter. In this manuscript, we show that the same principle can be applied to create non-equilibrium correlated states with spontaneously broken symmetry in a lightly doped semiconductor. The periodic drive provides means for obtaining large electronic densities of states necessary for the broken symmetry phase. The phase transition occurs in the steady-state of the system achieved due to interplay between the coherent external drive, electron-electron interactions, and dissipative processes arising from the coupling to phonons and the electromagnetic environment. We obtain the phase diagram of the system using numerical calculations that match predictions obtained from a phenomenological treatment and discuss the conditions on the system and the external drive under which spontaneous symmetry breaking occurs. Our results imply that Floquet engineering of the density of states provides a new route for generating and controlling correlated states of electrons with external fields.

cond-mat.str-el↗

arXiv2019

Higgs-Axion interplay and anomalous magnetic phase diagram in TlCuCl$_3$

What is so unique in TlCuCl3 which drives so many unique magnetic features in this compound? To study these properties, here we employ a combination of ab-initio band structure, tight-binding model, and an effective quantum field theory. Within a density-functional theory (DFT) calculation, we find an unexpected bulk Dirac cone without spin-orbit coupling (SOC). Tracing back to its origin, we identify, for the first time, the presence of a Su-Schrieffer-Heeger (SSH) like dimerized Cu chain lying in the 3D crystal structure. The SSH chain, combined with SOC, stipulates an anisotropic 3D Dirac cone where chiral and helical states are intertwined. As a Heisenberg interaction is introduced, we show that the dimerized Cu sublattices of the SSH chain condensate into spin-singlet, dimerized magnets. In the magnetic ground state, we also find a topological phase, distinguished by the axion angle. Finally, to study how the topological axion term couples to magnetic excitations, we derive a Chern-Simons-Ginzburg-Landau action from the 3D SSH Hamiltonian. We find that axion term provides an additional mass term to the Higgs mode, and a lifetime to paramagnons, which are independent of the quantum critical physics. The axion-Higgs interplay can be probed with electric and magnetic field applied parallel or anti-parallel to each other.

cond-mat.str-el↗

arXiv2018

Statistics-tuned phases of pseudofermions in one dimension

We show that a quadratic system of pseudofermions, with tunable fractionalised statistics, can host a rich phase diagram on a one dimensional chain with nearest and next nearest neighbor hopping. Using a combination of numerical and analytical techniques, we show that that by varying the statistical angle and the ratio of the hopping, the system stabilizes two Tomonaga-Luttinger liquids (TLL) with central charges c = 1 and 2 respectively along with the inversion symmetry broken bond ordered (BO) insulating phase. Interestingly, the two quantum phase transitions in the system - (1) between the two TLLs, and, (2) the c = 1 TLL and BO phase can be engendered by solely tuning the statistics of the pseudofermions. Our analysis shows that both these transition are continuous and novel with the former lacking a local order-parameter based description and the latter of Berezinskii-Kosterlitz-Thouless type. These phases and phase transitions can be of direct experimental relevance in context of recent studies of fermionic cold atoms.

cond-mat.str-el↗

arXiv2018

Surprises in the t-J model: Implications for cuprates

The t-J model is a paradigmatic model for the study of strongly correlated electron systems. In particular, it has been argued that it is an appropriate model to describe the cuprate high-Tc superconductors. It turns out that a comprehensive understanding of the gamut of physics encoded by the t-J model is still an open problem. In recent years some remarkable experiments on the cuprates, for example, discovery of nodeless superconductivity in underdoped samples (PNAS 109, 18332 (2012)), discovery of s-wave like gap in the pseudogap phase (Phys. Rev. Lett. 111, 107001 (2013)), and observation of polar Kerr effect (PKE) (Phys. Rev. Lett. 112, 047003 (2014)), have thrown up new challenges for this model. Here, we present results demonstrating that, within the slave-particle formulation of the t-J model, the d-wave superconductor is unstable at low doping to its own anti-symmetric phase mode fluctuations when the effect of fluctuations is treated self-consistently. We then show that this instability gives way to a time reversal symmetry broken d + is-SC in the underdoped region which has superfluid stiffness consistent with Uemura relation, even with a large pair amplitude. We show that our results are consistent with existing experiments on cuprates and suggest that Josephson (SQUID interferometry) experiments can clearly distinguish the d+is-SC from a host of other possibilities alluded to be contributing to the physics of underdoped cuprates. We also comment on other theoretical studies vis-a-vis ours.

cond-mat.supr-con↗

arXiv2018

Large Landau level splitting with tunable one-dimensional graphene superlattice probed by magneto capacitance measurements

The unique zero energy Landau Level of graphene has a particle-hole symmetry in the bulk, which is lifted at the boundary leading to a splitting into two chiral edge modes. It has long been theoretically predicted that the splitting of the zero-energy Landau level inside the {\it bulk} can lead to many interesting physics, such as quantum spin Hall effect, Dirac like singular points of the chiral edge modes, and others. However, so far the obtained splitting with high-magnetic field even on a hBN substrate are not amenable to experimental detection, and functionality. Guided by theoretical calculations, here we produce a large gap zero-energy Landau level splitting ($\sim$ 150 meV) with the usage of a one-dimensional (1D) superlattice potential. We have created tunable 1D superlattice in a hBN encapsulated graphene device using an array of metal gates with a period of $\sim$ 100 nm. The Landau level spectrum is visualized by measuring magneto capacitance spectroscopy. We monitor the splitting of the zeroth Landau level as a function of superlattice potential. The observed splitting energy is an order higher in magnitude compared to the previous studies of splitting due to the symmetry breaking in pristine graphene. The origin of such large Landau level spitting in 1D potential is explained with a degenerate perturbation theory. We find that owing to the periodic potential, the Landau level becomes dispersive, and acquires sharp peaks at the tunable band edges. Our study will pave the way to create the tunable 1D periodic structure for multi-functionalization and device application like graphene electronic circuits from appropriately engineered periodic patterns in near future.

cond-mat.mes-hall↗