https://dspraneeth07.in/researchResearch and publications by Dhadi Sai Praneeth Reddy.enhttps://dspraneeth07.in/research/neurogenesishttps://dspraneeth07.in/research/neurogenesisDhadi Sai Praneeth ReddyPreprintATLAS-R-2026-008. Compiler optimization passes are hand-written heuristics. Learned approaches such as MLGO and AutoPhase improve on them, but still choose among predefined passes. NeuroGenesis encodes compiler IR with a Transformer and trains a reinforcement-learning policy to discover optimization rules autonomously. A 0.28 reduction in execution cost relative to the LLVM baseline (0.00), outperforming MLGO and AutoPhase.https://dspraneeth07.in/research/kiranachainhttps://dspraneeth07.in/research/kiranachainDhadi Sai Praneeth ReddyDatasetdoi:10.57967/hf/9015ATLAS-R-2026-007. Agent negotiation benchmarks rarely reflect informal economies, where credit is extended on trust, supply is irregular and trade happens across languages. KiranaChain models India's kirana-store economy as a multi-agent negotiation environment covering informal credit, supply-chain dynamics and multilingual commerce. An openly released benchmark on Hugging Face (DOI 10.57967/hf/9015) for evaluating negotiating agents in a setting that matters to 12–14 million stores. DOI: 10.57967/hf/9015https://dspraneeth07.in/research/gct-100khttps://dspraneeth07.in/research/gct-100kDhadi Sai Praneeth ReddyDatasetATLAS-R-2026-006. Digital fraud works by manipulating people (fear, urgency, trust), yet most security datasets label messages only as malicious or benign. GCT-100K collects 100,000 samples across phishing, OTP fraud and social engineering, and introduces the Cognitive Manipulation Score (CMS) to quantify manipulation. An Apache 2.0 benchmark for measuring how manipulative a message is, not just whether it is malicious.https://dspraneeth07.in/research/cognitive-attack-topologyhttps://dspraneeth07.in/research/cognitive-attack-topologySun, 01 Mar 2026 09:00:00 GMTDhadi Sai Praneeth ReddyM. Jithender ReddyGNR PrasadPreprintdoi:10.5281/zenodo.19303021ATLAS-R-2026-004. The dominant paradigm in cybersecurity continues to privilege infrastructure, protocol integrity, and endpoint resilience, while an increasing fraction of high-impact attacks bypasses these controls by directly manipulating human cognition. This work formalizes such attacks as structured distortions within human digital trust interactions, introducing Cognitive Topological Cybersecurity (CTC) as a rigorous analytical framework. Within this paradigm, interactions are modeled as a high-dimensional manifold whose geometry encodes trust relationships across users, communication channels, and psychological signals. The proposed Cognitive Attack Topology (CAT) framework operationalizes this view through the construction of a Trust Topology Tensor (TTT), enabling the quantification of adversarial influence via Cognitive Distortion Energy (CDE) and its normalized form, the Trust Distortion Index (TDI). A complementary Cognitive Manipulation Score (CMS) captures the composite effect of urgency, fear, authority, and persuasion signals. The framework is instantiated in a multi-layer architecture that integrates transformer-based signal decomposition with dynamic graph modeling and topological anomaly detection. Empirical evaluation is conducted on the GCT-100K dataset, a large-scale benchmark comprising real, public, and synthetically generated cognitive attack interactions. The CAT model achieves a ROC-AUC of 0.9555, exceeding conventional text-based baselines, while maintaining robustness under adversarial cloaking conditions. Notably, performance remains invariant across linguistic boundaries, with a multilingual AUC of 0.984, indicating that topological features of trust manipulation exhibit language-agnostic structure. These results establish that human-targeted cyber-attacks can be detected not as isolated semantic artifacts, but as measurable geometric perturbations in trust space, motivating a shift toward cognition-aware, topology-driven defensive systems. DOI: 10.5281/zenodo.19303021https://dspraneeth07.in/research/aolphttps://dspraneeth07.in/research/aolpSun, 01 Feb 2026 09:00:00 GMTDhadi Sai Praneeth ReddyPreprintdoi:10.5281/zenodo.18641867ATLAS-R-2026-003. Current artificial intelligence research is dominated by the paradigms of optimization, learning, and evolution. While these frameworks have yielded high-performance task-specific agents, they operate under the implicit assumption that the space of agent development is unconstrained: intelligence is viewed as a fluid commodity that emerges proportionally to data volume, compute, or reward density. This paper argues that such an approach ignores a fundamental scientific dimension: the structural laws that govern how an agent is allowed to become intelligent over time. We introduce Agent Ontogeny & Lineage Physics (AOLP), a foundational framework that reframes artificial development as a constrained physical process governed by deterministic laws. By shifting the focus from "what an agent learns" to "the lawful constraints under which it develops," we define the Developmental Trajectory Constraint System (DTCS), a new scientific object that enforces irreversibility and phase-dependent growth. We formalize the primitives of Developmental Laws (DLs), Ontogenetic Phases, and Lineage-level constraints. Unlike traditional regularization or curriculum learning, AOLP treats these constraints as "physics": structural boundaries that, when violated, result in irreversible developmental regression rather than transient performance penalties. This paradigm provides the mathematical and conceptual tools to analyse Developmental Collapse and Counterfactual Recovery Horizons, offering a new axis for understanding AI safety, stability, and the transition toward robust, self-modelling artificial intelligences. DOI: 10.5281/zenodo.18641867https://dspraneeth07.in/research/solipur-agriculture-ai-ml-iothttps://dspraneeth07.in/research/solipur-agriculture-ai-ml-iotThu, 01 Jan 2026 09:00:00 GMTJalaja TattariT. AdilakshmiDhadi Sai Praneeth ReddyConference Paperdoi:10.1007/978-981-95-0144-1_16ATLAS-R-2026-002. The combination of Artificial Intelligence, Machine Learning, and the Internet of Things in farming is changing how farmers work. This paper discusses how Solipur Village uses these technologies to improve its farming methods. With AI predicting outcomes, ML managing crops, and IoT monitoring in real time, the study looks at better crop yield, resource use, and farmer earnings. Data from fieldwork, surveys, and interviews show the benefits and challenges of using these technologies. The results show higher productivity and sustainability, giving suggestions for policymakers and farmers to adopt AI, ML, and IoT in Indian agriculture. DOI: 10.1007/978-981-95-0144-1_16https://dspraneeth07.in/research/eagleaihttps://dspraneeth07.in/research/eagleaiThu, 01 Jan 2026 09:00:00 GMTDhadi Sai Praneeth ReddyTechnical ReportATLAS-R-2026-001. A proprietary language model built from scratch for Telangana's EAGLE Force: custom architecture, tokenizer and pretraining pipeline.https://dspraneeth07.in/research/redefining-agihttps://dspraneeth07.in/research/redefining-agiSat, 01 Nov 2025 09:00:00 GMTDhadi Sai Praneeth ReddyAmulya BiradarManikanth Reddy MaramShreya BurraPreprintdoi:10.13140/RG.2.2.20594.16324ATLAS-R-2025-005. Artificial General Intelligence (AGI) has remained largely theoretical due to vague definitions, non-measurable criteria, and architectures that cannot be implemented in practice. Existing interpretations of AGI, from cognitive theories to universal intelligence models, provide valuable insights but do not offer a concrete pathway for building or evaluating an actual general intelligence system. This paper introduces a new, measurable, and operational definition of AGI that emphasizes autonomous knowledge acquisition, reasoning across diverse and clearly defined domains, cross-domain transfer, adaptive self-improvement, and alignment with human goals. To support this definition, we propose a modular cognitive framework designed specifically for practical implementation. A working prototype is developed to demonstrate the feasibility of this approach. The system is capable of learning new knowledge, storing it in an adaptive memory, applying multi-step reasoning, transferring understanding across unrelated domains, and improving its performance through user feedback. Built using currently available technologies such as the Gemini API and structured memory mechanisms, the prototype shows that AGI can be demonstrated meaningfully even with today's tools. The paper also presents a standardized evaluation suite that measures generalization, transfer, reasoning accuracy, learning efficiency, memory retention, adaptability, and alignment stability. Together, the definition, architecture, and prototype form a complete foundation for practical AGI research and represent a significant step toward realizing general-purpose intelligence. DOI: 10.13140/RG.2.2.20594.16324https://dspraneeth07.in/research/aqi-prediction-hyderabadhttps://dspraneeth07.in/research/aqi-prediction-hyderabadSat, 01 Nov 2025 09:00:00 GMTDhadi Sai Praneeth ReddySoli Dhana SriSreenidhi PagidimarriMaram Manikanth ReddyJournal Articledoi:10.55248/gengpi.06.1125.3858ATLAS-R-2025-004. Especially in fast-expanding cities like Hyderabad, where industrial activity, traffic emissions, and weather conditions create difficult air pollution patterns, air quality forecasting is a serious issue in urban areas. Traditional approaches for predicting the Air Quality Index (AQI) sometimes fall short because they do not represent the non-linear relationships and time dependencies found in air quality information. The research proposes a multi-modal deep learning architecture using Long Short-Term Memory (LSTM), Transformer, and Random Forest models to study past air quality data from 15 monitoring stations in Hyderabad. With a Root Mean Square Error (RMSE) of 3.7, the framework beats conventional statistical techniques and one deep learning technique, with 94.65 percent accuracy. Among the most important contributions are the creation of a strong preprocessing pipeline, the identification of important features including PM2.5, temperature, and wind speed, and the validation of real-time processing capacity. Seasonal study shows major fluctuations, with winter months having PM2.5 levels 45% above the yearly average. In addition, the research emphasizes the practical effects of the framework for policy formulation and management of urban air quality by providing a scalable answer for real-time AQI forecasting and long-term trend analysis. DOI: 10.55248/gengpi.06.1125.3858https://dspraneeth07.in/research/decline-of-blockchain-popularityhttps://dspraneeth07.in/research/decline-of-blockchain-popularityMon, 28 Jul 2025 09:00:00 GMTDhadi Sai Praneeth ReddyKasireddy Manideep ReddyPreprintdoi:10.36227/techrxiv.175372475.58978266/v1ATLAS-R-2025-003. Blockchain technology, once considered revolutionary, has experienced a significant downturn in popularity over the past several months. The primary reasons behind this decline include technical challenges, regulatory uncertainty, and diminishing trust in the scalability and security of existing blockchain systems. This paper examines the multifaceted causes of blockchain's downward trajectory and presents a critical analysis of the current state of blockchain technologies, addressing specific technical bottlenecks such as energy inefficiency, governance issues, and lack of interoperability. The analysis is supplemented with real-time data, tables, graphs, and an exploration of emerging trends that could shape blockchain's future. DOI: 10.36227/techrxiv.175372475.58978266/v1https://dspraneeth07.in/research/5g-beyond-european-policyhttps://dspraneeth07.in/research/5g-beyond-european-policyTue, 25 Feb 2025 09:00:00 GMTJalaja TattariT. AdilakshmiDhadi Sai Praneeth ReddyP. Adithya ReddyConference Paperdoi:10.1007/978-981-97-9266-5_11ATLAS-R-2025-002. The emphasis of this paper is on the need to update regulations to promote the advancement of 5G and 6G technology in Europe, which is important for the future of the digital economy. The European Union needs a more effective strategy to lead the way in 5G technology, given its current rollout and significance. The focus is on devising a blueprint for the industry that can contribute to the advancement of the telecommunications sector in Europe. This plan would involve all countries in the European Union, rather than focusing on each country separately. DOI: 10.1007/978-981-97-9266-5_11https://dspraneeth07.in/research/multi-tenancy-fpga-cloudhttps://dspraneeth07.in/research/multi-tenancy-fpga-cloudWed, 01 Jan 2025 09:00:00 GMTBusa RevathiDhadi Sai Praneeth ReddyKasireddy Manideep ReddyJournal ArticleATLAS-R-2025-001. Multi-tenant FPGA cloud systems represent a new frontier in high-performance computing (HPC), blending flexibility, processing power, and cost efficiency. FPGA-based systems, integrated into cloud environments, hold immense potential for industries requiring real-time data processing, such as artificial intelligence (AI), finance, and healthcare. These systems enable dynamic resource allocation, offering better performance while addressing cost concerns. This paper presents a detailed exploration of the system architecture, challenges, implementation, and the impact of such systems, along with an evaluation of their suitability in line with India's government initiatives. Potential future research directions such as quantum-FPGA integration are also discussed.https://dspraneeth07.in/research/microseism-noise-indian-oceanhttps://dspraneeth07.in/research/microseism-noise-indian-oceanFri, 01 Nov 2024 09:00:00 GMTDhadi Sai Praneeth ReddyKasireddy Manideep ReddySahasra BaggariTanishka KoraConference Paperdoi:10.5281/zenodo.14102511ATLAS-R-2024-003. Within a span of 2 to 20 seconds, microseism noise occurs as a distinctive sound. It is the loudest band of sound in the earth's background noise. In this study, we looked at the size and direction of the small movements in the Indian Ocean. Data is gathered from a network of ten land stations located at various points across the Indian Ocean. The description of microseism was achieved by utilizing the probability power spectral density. To understand the microseism, we use a method that looks at how the vibration changes with different frequencies. This method uses a mathematical calculation called Eigenvalue decomposition to analyze the way the vibrations spread in three different directions. We studied how microseism changes over space and time in the Indian Ocean area. The Indian Ocean is considered a major contributor to microseism worldwide. In certain stations, we can see that the double frequency microseism breaks down into short periods of 2 to 5 seconds and long periods of 6 to 10 seconds. The study shows that the main causes of the continuous low-frequency seismic vibrations come from the Southern Ocean. In addition, we conducted a comparison between the alterations in wave heights and the intensity of waves at each particular site. It was evident that sound wave intensity and wave height within a particular frequency range were closely linked. This connection was observed both in different places and at different times. We additionally include an overview of the prevalent wave types detected in the microseism band on the Earth's surface. Over an extended duration, Rayleigh waves are the prevalent ones, whereas Love waves tend to be more prominent in shorter intervals. DOI: 10.5281/zenodo.14102511https://dspraneeth07.in/research/generative-data-augmentation-cganshttps://dspraneeth07.in/research/generative-data-augmentation-cgansFri, 01 Nov 2024 09:00:00 GMTRevathi BusaDhadi Sai Praneeth ReddyKasireddy Manideep ReddySahasra BaggariConference Paperdoi:10.5281/zenodo.14163827ATLAS-R-2024-002. Imbalanced datasets present a significant challenge in machine learning, particularly in tasks such as fraud detection, medical diagnosis, and sentiment analysis. This paper proposes a novel approach using Conditional Generative Adversarial Networks (CGANs) to generate synthetic samples that enhance the performance of models trained on imbalanced datasets. By conditioning the generation process on class labels, CGANs produce realistic and diverse samples of the underrepresented classes. The proposed method is empirically evaluated on various benchmark datasets, demonstrating superior performance in comparison to traditional oversampling methods like SMOTE (Synthetic Minority Over-sampling Technique). Our results show that CGANs significantly improve the predictive accuracy, precision, recall, and F1-score, leading to more balanced and reliable machine learning models. DOI: 10.5281/zenodo.14163827https://dspraneeth07.in/research/cybersecurity-insights-world-religionshttps://dspraneeth07.in/research/cybersecurity-insights-world-religionsFri, 01 Nov 2024 09:00:00 GMTDhadi Sai Praneeth ReddyReview Paperdoi:10.5281/zenodo.14061606ATLAS-R-2024-001. To secure firm information, organizations develop and communicate security guidelines for employees, yet employees frequently don't follow them because of human characteristics. Experts advise taking lessons from other disciplines that deal with irrational human behavior, such as religions, in order to overcome this problem. These disciplines have dealt with human errors and flaws for thousands of years. Cybersecurity experts may gain significant insight from faiths, which have evolved strategies for assisting the weak and safeguarding their adherents. A great amount of research was done to develop a cybersecurity plan, including examining religious literature and speaking with religious experts. The objective is to start conversations so that individuals can collaboratively think and find better ways to implement cybersecurity. This strategy can help cybersecurity experts create more efficient plans. DOI: 10.5281/zenodo.14061606