Digital Commons Network^™

Measuring The Tokenization Premium: A Cost Audit For Underserved Language Communities, Avijit Roy, Proma Roy, Hrishitva Patel

Publications and Research

Large language models are increasingly deployed as general-purpose educational and technical assistance systems, but their basic infrastructure does not treat languages equally. One underexamined source of disparity is tokenization: semantically equivalent content can require substantially different token counts across languages, affecting API cost, latency, and usable context length before a model is even invoked. We introduce the Tokenization Equity Audit (TEA), a reproducible benchmark for measuring tokenization premiums in technical tutoring content. TEA evaluates three widely used tokenizers, GPT-4o’s o200k base, Qwen2.5-7B, and Mistral-7B, on a 120-item Python debugging corpus translated from English into Bengali, Hindi, Arabic, Tamil, and Yoruba. …

Effective Visibility In The Infrared Bands, Peter L. Dean-Erlander, Steven T. Fiorino, Ronald G. Driggers

Faculty Publications

Visibility is an atmospheric metric for the comparison of terrestrial imaging conditions and locations. A limitation of visibility is that it is defined for the visible spectrum only, and there is no simple infrared equivalent. This study compares three reflective infrared wavebands: near-IR (NIR), shortwave IR (SWIR), and extended shortwave IR (eSWIR), to the visible band across a global set of cities to develop three rule-of-thumb functions for IR effective visibility. To accomplish this, a radiometric sensor model is combined with the Laser Environmental Effects Definition and Reference (LEEDR) software package to calculate the visibility of a black-and-white contrast target …

The Mathematics Of Calendar Patterns: Day-Of-Week Phenomena, Number Patterns, And Recreational Investigations, Vijender Nallapu, Mrinal Kanti Roychowdhury

School of Mathematical & Statistical Sciences Faculty Publications

Calendars are among the most familiar mathematical structures encountered in everyday life. Beyond their practical use in organizing dates and measuring time, calendars contain a rich collection of numerical patterns, symmetries, and arithmetic relationships. Many of these patterns can be explored using only elementary mathematics, making calendars an attractive subject for recreational investigation and mathematical enrichment.

In this article we examine several mathematical properties arising from the structure of monthly calendars. We investigate arithmetic patterns associated with rows, columns, and rectangular blocks of dates, establish properties of calendar squares, and explore connections with modular arithmetic through day-of-week calculations. We also …

Vedic Mathematics And Mental Computation: Fast Multiplication Techniques, Algebraic Foundations, And Historical Perspectives, Rakesh Bhatia, Mrinal Kanti Roychowdhury

School of Mathematical & Statistical Sciences Faculty Publications

Mental computation has fascinated mathematicians, educators, and students for centuries. Various computational techniques allow arithmetic calculations to be performed rapidly without the use of calculators or computers. Among these methods, several popular techniques are associated with what is commonly known as Vedic Mathematics. In this article, we examine a collection of fast multiplication techniques, including multiplication near powers of ten, multiplication by eleven, squaring numbers ending in five, and the vertical-and-crosswise method. For each technique, we provide elementary mathematical explanations based on algebraic identities and place-value representations.

Hidden Symmetries Of The Classical 3×3 Magic Square, Vijender Nallapu, Mrinal Kanti Roychowdhury

School of Mathematical & Statistical Sciences Faculty Publications

The classical 3×3 magic square is one of the most famous objects in recreational mathematics. Its defining feature is that every row, every column, and the two main diagonals have the same sum. However, beyond this well-known property, the square contains several hidden symmetries involving rotations, reflections, complementary pairs, parity, and balance around the central entry. In this article, we present an elementary exploration of these symmetries and show how the classical magic square provides a beautiful introduction to mathematical structure and pattern recognition.

Curious Patterns And Identities Involving Consecutive Integers, Carlo Cattani, Mrinal Kanti Roychowdhury

School of Mathematical & Statistical Sciences Faculty Publications

Consecutive integers exhibit numerous surprising patterns and elegant identities. In this article, we investigate several classical properties involving products, sums, divisibility, perfect squares, cubes, pronic numbers, triangular numbers, sums of squares, and Pythagorean triples. Using only elementary algebra, symmetry arguments, and mathematical induction, we establish a collection of identities that reveal rich connections among these classical topics. The exposition is intended for students, teachers, and recreational mathematicians, illustrating how simple observations about consecutive integers can lead to beautiful mathematical patterns and deeper insights into elementary number theory.

Gwtc-4.0: Constraints On The Cosmic Expansion Rate And Modified Gravitational-Wave Propagation, A. G. Abac, I. Abouelfettouh, Teviet Creighton, Mario C. Diaz, Raul Alberto Espinosa Perez, J. Lawrence, Francisco Llamas Villarreal, Soma Mukherjee, Volker Quetschke, Miriam Ramos Arevalo, Wenhui Wang

Physics & Astronomy Faculty Publications

We analyze data from 142 of the 218 gravitational-wave (GW) sources in the fourth LIGO-Virgo-KAGRA Collaboration (LVK) Gravitational-Wave Transient Catalog (GWTC-4.0) to estimate the Hubble constant H0 jointly with the population properties of merging compact binaries. We measure the luminosity distance and redshifted masses of GW sources directly; in contrast, we infer GW source redshifts statistically through (i) location of features in the compact object mass spectrum and merger rate evolution, and (ii) identifying potential host galaxies in the GW localization volume. Probing the relationship between source luminosity distances and redshifts obtained in this way yields constraints on cosmological parameters. …

Moss Biomonitoring And Isotopic Signatures Reveal Aviation Fuel As Local Lead Source Near Oregon Airports, Oluwaseyi E. Aderinboye, Alyssa E. Shiel, Lauren Stalford, Christina J. Murphy, Dean B. Atkinson

Chemistry Faculty Publications and Presentations

Lead emissions from piston-engine aircraft using leaded aviation gasoline (avgas) remain a significant source of environmental contamination near regional airports, posing potential health risks to nearby communities. Passive biomonitoring using the epiphytic moss Orthotrichum lyellii was combined with lead isotope fingerprinting to quantify lead concentrations and distinguish between legacy leaded gasoline and current avgas emissions near Hillsboro and Troutdale airports in Oregon. Here we show that lead levels are elevated by up to 50 times the rural background near airports, with isotopic signatures confirming avgas as a dominant local source. These findings indicate that avgas use contributes substantially to environmental …

Building Agentic Ai Systems With Langgraph: From Llm Pipelines To Multi-Agent Decision Workflows, Mohammad Amin Kuhail

All Works

Recent advances in large language models (LLMs) have enabled the development of increasingly sophisticated agentic AI systems capable of reasoning, planning, tool use, memory management, and collaborative problem solving. Agentic AI refers to AI systems that can autonomously pursue goals and execute tasks by combining reasoning, decision-making, memory, interaction with external tools, and adaptive workflows. Unlike traditional LLM applications that primarily rely on single-prompt interactions or fixed sequential pipelines, agentic AI systems can iteratively analyze problems, coordinate multiple steps, invoke external resources, refine intermediate outputs, and dynamically adapt their behavior in response to evolving context and feedback. Recent developments in …

Adaptive Phishing Url Detection Using Hybrid Fuzzy C-Means Clustering And Xgboost, Muntadher Mohammed Kareem, Rawaa I. Farhan

Karbala International Journal of Modern Science

Phishing attacks continue to evolve in sophistication, rendering static detection methods increasingly ineffective. Existing URL-based approaches suffer from limited adaptability to emerging phishing patterns, mislabeled training data, and insufficient validation protocols. This paper proposes a hybrid phishing URL detection system that integrates Fuzzy C-Means (FCM) clustering with XGBoost classification, enhanced by a novel Micro Adaptive Feature Extractor (MAFE). The system employs a multi-stage pipeline: feature engineering generating 36 statistical and interaction features, MAFE producing 15 adaptive features through class-aware dynamic weighting, micro-pattern detection, and entropy analysis, and FCM with K=2 clusters providing soft membership features to XGBoost. A two-pass confidence-based …

Learning, Abstraction, And Creative Search For Interactive Agents, Jonathan Demke

Theses and Dissertations

Artificial intelligence systems are increasingly deployed as autonomous agents that interact directly with complex environments. Ensuring these agents remain reliable, adaptable, and useful presents an ongoing practical challenge. This dissertation studies agent-environment interactions through a progressive lens, examining behavior from foundational training stability and abstraction discovery to adaptation in unpredictable settings and open-ended creative search. The investigation begins by analyzing how hyperparameter settings affect early-stage Deep Q-Network performance, yielding practical guidelines for debugging reinforcement learning implementations. To address the challenge of long-horizon planning, the discovery of temporally extended options is formalized as the Minimum Shortcut Problem, demonstrating how agents can …

Development Of Minimalist Trifunctional Building Blocks For Chemical Biology Probe Synthesis, Luke E. Bohrer

Electronic Theses and Dissertations

The completion of the Human Genome Project in 2003 generated expectations that genomic information would accelerate the discovery of new protein drug targets and FDA-approved therapies. However, many small-molecule drug candidates continue to fail in clinical trials, and much of the human proteome remains undercharacterized in terms of structure, function, and disease relevance. Chemical tools are therefore essential for defining protein function, mapping protein interactions, and identifying disease-associated molecular mechanisms. Among these tools, small-molecule chemical probes provide a powerful strategy for connecting biological activity with target identification, functional analysis, and the development of effective therapies.

This dissertation describes the development …

Wip: Developing A Generative Ai Autoethnography Assistant, Jyoti Suhag, Jennifer Drewyor, Kathryn Bugbee, Michelle Jarvie-Eggart, Lynn Albers, Leo Ureel

Michigan Tech Publications

Background: Generative AI (genAI) is transforming educational research, offering new possi-bilities for conducting interviews while local sandboxing minimizes data privacy risks and hallucinations. Purpose: This work-in-progress presents the AI Autoethnography Assistant, a project investigating how large language models (LLMs) can support autoethnographic interview design and execution. Approach: Using prompt engineering grounded in Interpretative Phenomenological Anal-ysis, paraphrasing techniques, and structured follow-up questions, we developed protocols that incorporate po-sitionality and prompt reflection on origin stories and pivotal life moments. Outcomes: Conversations conclude at the user’s discretion, with the AI generating a thematic summary. We tested refined prompts across four platforms (Gemini, Claude, …

Fragberta: A Fragment-Aware Molecular Representation Model With Sequential Attachment-Based Fragment Embeddings, Neerav Kaushal, Ajay M.N.V. Penmatsa

Michigan Tech Publications

Automated exploration of chemical space through machine learning has become a central strategy in modern drug discovery, where the ability to learn transferable molecular representations directly impacts predictive modeling performance. Recent advances in chemical language models have primarily relied on atom-level string encodings such as SMILES or SELFIES, treating molecules as sequences of individual atoms or characters. However, medicinal chemistry reasoning is often fragment-centric, emphasizing scaffolds, functional groups, and modular substructures rather than isolated atoms. In this study, we introduce FragBERTa, a molecular fragment-aware transformer-based representation language model pretrained using masked language modeling on Sequential Attachment-based Fragment Embedding (SAFE) representations. …

Modeling And Simulation Of Solar Battery Charge Controller Using Adaptive Particle Swarm Optimization Mppt Algorithm, Mustafa Sacid Endiz, Göksel Gökkus

Mathematical Modelling and Numerical Simulation with Applications

Implementing an effective Maximum Power Point Tracking method is crucial for optimizing solar energy harvesting against environmental fluctuations like solar radiation and temperature. This paper introduces a novel approach for modeling and simulating a solar battery charge controller, using a modified Particle Swarm Optimization algorithm. The power stage of the system is based on a SEPIC converter, which is employed to manage the power conversion and improve the energy transfer to the battery. The developed circuit model is evaluated under various radiation levels at a constant temperature, as well as under different temperature levels at a constant radiation. Simulations are …

Technical Reliability As Foundation Of Mobile Self-Checkout Experience, Pumezile O. Dyantyi, Fazlyn Petersen

African Conference on Information Systems and Technology

Although mobile self-checkout (MSCO) apps are becoming more common in retail stores, their long-term success in emerging markets depends heavily on the system’s technical reliability. This qualitative thematic study of 888 user reviews for four established MSCO apps investigates how technical reliability shapes the user experience (UX) after app adoption. Using an integration of the UX Honeycomb model and the UTAUT2 framework, this paper demonstrates that technical reliability is a core UX element that affects usability, performance, trust, value, satisfaction and continued use. Technical reliability problems such as system outages, login challenges and transaction difficulties negatively impact several core UX …

South Africa's Inclusive Six-Dimensional Generative Ai Readiness Framework, Keegan Steyn, Ojelanki Ngwenyama, Sibongiseni Tunzelana

African Conference on Information Systems and Technology

Generative artificial intelligence (Gen AI) is transforming governance, knowledge systems, and economic activity, but current preparedness measures pay little attention to the socio-technical, institutional, and postcolonial contexts that shape country trajectories. To address this gap, the study employs a design science methodology to create the National Generative AI Readiness Assessment Framework (N-GAIRF), a six-dimensional diagnostic tool theoretically based on Sen's capability approach to development and incorporating Institutional Theory, Socio-Technical Systems Theory, National Innovation Systems Theory, and Postcolonial Information Systems critique. An example application to South Africa, compared to Zimbabwe and an indicative Organization for Economic Cooperation and Development (OECD) reference …

Secondary Water Saturation Of The Rock Mass And Quality Of Rocks In Active Mine Areas Bordering The Flooded Workings Of Closed Mines, Mirosława Bukowska

Journal of Sustainable Mining

The goal of this paper was to inspect the influence of capillary water saturation on the specific strain energy of sedimentary rocks. The relationships between the determined specific strain energies in an air-dried state and in a state of capillary water saturation for detrital rock were exponential, with a strong correlation among the variables. For silt, the relationships were described by a second-degree polynomial function, and the correlation was medium or weak. The relationships for kinetic and dissipated energy were described with a linear function. Pearson correlation coefficients indicated a strong correlation among the variables for detrital rock and a …

The Role Of Trauma-Informed Care In Supporting At-Risk Students In Alternative Education, Jar'mel D. Taylor Dr.

Transformations

Students with traumatic experiences commonly fail to meet educational goals and control their emotions while also losing their involvement in classes. Standard trauma-informed care practices remain absent from numerous alternative educational institutions, which leaves their teaching community without adequate support to help traumatized students effectively. The application of Trauma-Informed Care (TIC) strategies toward student outcomes requires investigation to develop effective educational approaches that enhance school success. The researcher used qualitative exploratory case study methodology to study TIC practices in alternative education for Black and Brown adolescent students (aged 12–18) in Washington, D.C. Semi-structured interviews and a focus group with 16 …

Process From Noise: Erosion Threshold Variability And The Morphology Of Gravel-Bed Rivers, Robert Paul Kostynick

Arts & Sciences Graduate Student Theses and Dissertations

Rivers are remarkably consistent as their geometry, width, depth, and slope change systematically downstream in response to differences in fluid discharge and sediment transport throughout a channel. Rivers move both water and sediment through terrains that depend on these downstream geometric adjustments influencing water quality, ecology, infrastructure, channel morphology, and landscape evolution at increasing timescales respectively. Rivers thus influence all aspects of life and define our spatial landscapes that shape boundaries, whether physical, politically, or socially. Nevertheless, a river is dynamic and does not adhere to human-defined boundaries, making it vital that we understand how these features change over time …

Modeling Formation Of Turbulent Sporadic-E Clouds Using Realistic Wind Data, Aaron M. Schinder, Kenneth S. Obenberger, Jorge L. Chau, Juan M. Urco, Matthias Clahsen, Benjamin F. Akers, Daniel J. Emmons

Faculty Publications

A high resolution two-dimensional multi-fluid model of sporadic-E layers was developed and driven with physically realistic mesosphere, lower thermosphere (MLT) winds measured over Albuquerque, New Mexico. The realistic E-region winds are produced by the HYdrodynamic Point-wise Environment Reconstructor (HYPER) model that ingests meteor derived wind observations from a Spread-spectrum Interferometric Multistatic meteor radar Observing Network (SIMONe) system combined with the Navier-Stokes equations to provide high resolution three-dimensional wind fields over time. Sporadic-E dynamics are simulated using both realistic winds from HYPER as well as idealized hyperbolic tangent windshears to compare and contrast. Overall, the model shows greater inhomogeneity and irregularity …

Gwtc-4.0: Constraints On The Cosmic Expansion Rate And Modified Gravitational-Wave Propagation, The Ligo Scientific Collaboration, The Virgo Collaboration, The Kagra Collaboration, Marc Favata, Shoan Ghosh, Rodica M. Martin

Department of Physics and Astronomy Faculty Scholarship and Creative Works

We analyze data from 142 of the 218 gravitational-wave (GW) sources in the fourth LIGO–Virgo–KAGRA Collaboration (LVK) Gravitational-Wave Transient Catalog (GWTC-4.0) to estimate the Hubble constant H0 jointly with the population properties of merging compact binaries. We measure the luminosity distance and redshifted masses of GW sources directly; in contrast, we infer GW source redshifts statistically through (i) location of features in the compact object mass spectrum and merger rate evolution, and (ii) identifying potential host galaxies in the GW localization volume. Probing the relationship between source luminosity distances and redshifts obtained in this way yields constraints on cosmological parameters. …

Assessing The Impact Dynamics Of Less Lethal Ammunition, Steven Bermejo

Student Theses

Less-lethal ammunition is widely used by law enforcement as an alternative to lethal force during crowd control, on-street confrontations, and other situations requiring force. Although these projectiles are intended to reduce risk of fatal injury, serious injuries and fatalities have been reported, highlighting the need to better understand the factors that influence their impact behavior. This study evaluated the impact dynamics of four types of 12-gauge less-lethal ammunition, including rubber buckshot, single rubber ball, double rubber ball, and Byrna kinetic projectile, using kinetic energy calculations derived from projectile mass and velocity measurements. In addition, Attenuated Total Reflectance Fourier Transform Infrared …

Characterizing Candidate Blazar Counterparts Of The Ultra-High-Energy Event Km3-230213a, O. Adriani, S. Aiello, A. Albert, A. R. Alhebsi, M. Alshamsi, S. Alves Garre, S. Rani, Et Al.

Michigan Tech Publications

High-energy astrophysical neutrinos serve as crucial messengers for understanding hadronic acceleration processes and identifying the origins of cosmic rays, with blazars among the most promising neutrino sources. The KM3NeT experiment reported the detection of an ultrahigh-energy neutrino with an energy estimate of ∼220 PeV, the most energetic yet observed. The neutrino arrival direction has a 99% confidence region of 3° radius centered at R.A. 94 (Formula presented) .° 3, decl. −7 (Formula presented) .° 8 (J2000). In this work, 17 candidate blazars located within this region are identified. Comprehensive new observations and archival data analysis for these sources are presented. …

Dry After-Ripening Increases Germination Of Critically Imperilled Physaria Thamnophila (Zapata Bladderpod), Christopher A. Gabler, Jerald T. Garrett

School of Earth, Environmental, & Marine Sciences Faculty Publications

Effective endangered plant conservation and recovery require empirical knowledge of seed dormancy release and responses to post-collection handling, yet such information is lacking for many rare taxa. Physaria thamnophila (Zapata bladderpod; Brassicaceae), a critically imperilled (G1) species endemic to southern Texas, USA, is poorly studied, and basic questions regarding seed storage and germination remain unresolved. We evaluated the effects of short-term cool-storage duration (14–180 d), storage temperature (refrigeration vs. freezing), dry after-ripening and maternal plant identity on germination probability and timing in P. thamnophila. Dry after-ripening and increased cool-storage duration consistently increased germination and shortened germination time, whereas storage temperature …

The Computed Microwave Spectrum Of The Protonated Fullerene C60H+, Laszlo Nemes, Jos Oomens, Vincent J. Esposito, Vincent Boudon, Alexander G. G. M. Tielens

Biology, Chemistry, and Environmental Sciences Faculty Articles and Research

The largest known molecule in the interstellar medium of galaxies, C₆₀, has been detected in its neutral and cationic form through its vibrational, UV-driven fluorescence emission spectrum and its electronic absorption spectrum, respectively. The detection of several polycyclic aromatic hydrocarbon molecules through their pure rotation spectrum in cold, dense, molecular cloud cores suggests that C₆₀ might be present in these environments as well. The low flux of UV pumping photons in molecular cloud cores and the absence of suitably bright background stars, make detection of C₆₀ and its cation through the commonly used methods impractical. As …

A Granularity-Centered Taxonomy Of Personalized Federated Learning, Ei Ei Nyein Chan, Sergei Chuprov, Pretom Roy Ovi, Kamrul Hasan

Computer Science Faculty Publications

Personalized Federated Learning (PFL) has emerged as a key approach to address performance degradation in FL systems under heterogeneous client data. While existing surveys typically categorize PFL methods based on optimization strategies or system-level mechanisms, they often overlook a fundamental question: where is personalization embedded within the model architecture? In this survey, we bridge this knowledge gap and introduce a granularity-centered taxonomy that organizes PFL approaches according to the structural depth of personalization, ranging from head-layer and layer-wise adaptation to model-wise and parameter-wise customization. This novel perspective helps practitioners select appropriate personalization strategies based on model architecture, data heterogeneity, and …

From Reluctance To Resilience: The Link Between Attitude And Growth Mindset In Calculus Students, Olivia Valente, Kathryn E. Pedings-Behling, Amy N. Langville

Journal of Educational Research and Practice

Many postsecondary students enter general education mathematics courses with negative attitudes, shaping their willingness to engage with the content. This study explores the relationship between students’ attitudes towards mathematics and their growth mindset, revealing effective strategies to enhance learning experiences, particularly for students who are reluctant learners. The research investigates two questions: (1) How do students’ Attitude Toward Mathematics Inventory (ATMI) scores and Growth Mindset Scale (GMS) scores relate? and (2) Do gender or course modality affect this relationship? This study addresses a gap in literature by examining how enjoyment, motivation, self-confidence, and perceived value connect with growth mindset. The …

Single Fluorogens And Orientation-Localization Microscopy For Quantifying Chemical And Biomolecular Dynamics At The Nanoscale, Yiyang Chen, Yuanxin Qiu, Matthew D. Lew

Electrical & Systems Engineering Publications and Presentations

Many chemical systems look uniform only because ensemble measurements average over their most interesting molecules. Electron-transfer rates vary across electrode surfaces; lipid membranes contain nanodomains with distinct packing and fluidity; peptide aggregates exhibit local polymorphism; and biomolecular condensates contain transient networks of interactions that are blurred in ensemble images. A central challenge in chemical imaging is not simply to see smaller structures, but to measure chemical variables such as polarity, redox potential, and molecular confinement at the single-molecule level. In this Account, we describe how fluorogens, molecules whose brightness, blinking, spectral shifts, orientation, rotational mobility, and motion are directly shaped …

Ammonium Phosphate Fertilizers Are A More Effective Phosphorus Source Than Superphosphate For Wheat Grown On Acidic Gravelly Soils, Craig A. Scanlan, Raj Malik, Fien Degryse, Andrew Van Burgel, Mike J. Mclaughlin

Grain and Other Field Crops Research Articles

Selecting the most effective phosphorus (P) fertilizer source is essential for maximizing crop yields. However, there is little information to guide the choice between the widely used soluble granular sources: ammonium phosphates and superphosphates. The aims of this study were to examine whether there are differences in wheat (Triticum aestivum L.) grain yield response to ammonium phosphate and superphosphate fertilizers in field conditions, and if these differences relate to diffusion of P from fertilizer granules. Two field experiments were conducted on acidic gravelly soils in Western Australia where P deficiency occurs to compare wheat growth response to P applied …