arXiv subjects

Michael Wagner

Publications and source records attributed to Michael Wagner.

At least 19 recordsLinked to original sources

arXiv2026

Keeping Models and Code in Sync: Roundtrip Engineering for Tactical Domain-Driven Design

Domain-Driven Design gives teams a shared vocabulary for complex business logic, but that vocabulary only stays useful as long as the model and the code agree with each other. In practice, they drift apart: code changes outpace the model, or model revisions never make it into the codebase. This paper presents JDomInO, a bidirectional synchronization toolchain for tactical DDD that keeps a Java codebase and its domain model connected through a shared metamodel, with the goal of keeping the two in sync as the system evolves. JDomInO generates Java code structure deterministically from a domain model (forward path) and reconstructs a domain model from existing Java code (reverse path). The forward path has been fully validated on a Hotel Management scenario covering all 12 building block types in the metamodel; the reverse path's mapping logic has passed unit testing, with end-to-end validation underway. We also outline how the structured domain model produced by JDomInO could serve as a precision context layer for AI code assistants, helping them respect aggregate boundaries and DDD semantics that raw source code alone does not convey.

cs.SE↗

arXiv2026

Montreal Forced Aligner and the state of speech-to-text alignment in 2026

The Montreal Forced Aligner (MFA) was released in 2016 and has since become the most widely used tool for forced alignment in research and industry. In the decade since, MFA has undergone substantial development, including expanded coverage across more languages and dialects using larger open-source datasets, harmonized IPA dictionaries, model adaptation, cross-language phone remapping, and support utilities. This paper documents MFA 3.0's developments since version 1.0 and evaluates MFA's performance across English, Japanese, and Korean, benchmarked against classic and neural forced aligners. MFA 3.0 achieves state-of-the-art or near state-of-the-art performance across all four benchmark datasets with mean boundary errors below 15 ms. Adaptation and cross-language remapping are effective for languages outside MFA's training distribution, and pronunciation probability modeling and phonological rules provide gains in specific conditions.

cs.CL↗

arXiv2026

Can LLMs Understand the Impact of Trauma? Costs and Benefits of LLMs Coding the Interviews of Firearm Violence Survivors

Firearm violence is a pressing public health issue, yet research into survivors' lived experiences remains underfunded and difficult to scale. Qualitative research, including in-depth interviews, is a valuable tool for understanding the personal and societal consequences of community firearm violence and designing effective interventions. However, manually analyzing these narratives through thematic analysis and inductive coding is time-consuming and labor-intensive. Recent advancements in large language models (LLMs) have opened the door to automating this process, though concerns remain about whether these models can accurately and ethically capture the experiences of vulnerable populations. In this study, we assess the use of open-source LLMs to inductively code interviews with 21 Black men who have survived community firearm violence. Our results demonstrate that while some configurations of LLMs can identify important codes, overall relevance remains low and is highly sensitive to data processing. Furthermore, LLM guardrails lead to substantial narrative erasure. These findings highlight both the potential and limitations of LLM-assisted qualitative coding and underscore the ethical challenges of applying AI in research involving marginalized communities.

cs.CL↗

arXiv2026

Beyond Solving: Prescriptive Probing for Neural Routing Solvers

Neural combinatorial optimization (NCO) trains fast heuristics for routing problems, but planners often need more than a single solve: they ask which stop to drop, which transition to preserve, or which subset of stops to remove if a route is infeasible. Answering such counterfactual questions by re-solving each candidate is expensive even when the action set is small. We introduce Prescriptive Probing: using the frozen representations of a trained NCO model to rank candidate interventions for what-if decision support. While recent interpretability work descriptively probes what NCO solvers encode, we ask whether those same representations can prescriptively guide intervention choices on real-road routing problems. Where exhaustive labels are available, we train supervised probes from offline re-solve labels; where they are not (e.g., combinatorial action spaces), we train a lightweight reinforcement learning intervention head. Across real-road benchmarks on the Asymmetric Traveling Salesperson Problem and Capacitated Vehicle Routing Problem, probes built on frozen model representations achieve the strongest performance on a majority of intervention tasks and improve over local-repair heuristics on edge-forbiddance and on key node- and multi-node-removal metrics. We further investigate probeability as a property of the routing model itself, varying model quality, the mix of supervised learning and reinforcement learning, and the architecture family. Overall, our results suggest a new use case for NCO solvers: representations learned for route construction can transfer to decision support.

cs.LG↗

arXiv2025

Mechanistic Interpretability for Neural TSP Solvers

Neural networks have advanced combinatorial optimization, with Transformer-based solvers achieving near-optimal solutions on the Traveling Salesman Problem (TSP) in milliseconds. However, these models operate as black boxes, providing no insight into the geometric patterns they learn or the heuristics they employ during tour construction. We address this opacity by applying sparse autoencoders (SAEs), a mechanistic interpretability technique, to a Transformer-based TSP solver, representing the first application of activation-based interpretability methods to operations research models. We train a pointer network with reinforcement learning on 100-node instances, then fit an SAE to the encoder's residual stream to discover an overcomplete dictionary of interpretable features. Our analysis reveals that the solver naturally develops features mirroring fundamental TSP concepts: boundary detectors that activate on convex-hull nodes, cluster-sensitive features responding to locally dense regions, and separator features encoding geometric partitions. These findings provide the first model-internal account of what neural TSP solvers compute before node selection, demonstrate that geometric structure emerges without explicit supervision, and suggest pathways toward transparent hybrid systems that combine neural efficiency with algorithmic interpretability. Interactive feature explorer: https://reubennarad.github.io/TSP_interp

cs.LG↗

arXiv2025

Safety Factories - a Manifesto

Modern cyber-physical systems are operated by complex software that increasingly takes over safety-critical functions. Software enables rapid iterations and continuous delivery of new functionality that meets the ever-changing expectations of users. As high-speed development requires discipline, rigor, and automation, software factories are used. These entail methods and tools used for software development, such as build systems and pipelines. To keep up with the rapid evolution of software, we need to bridge the disconnect in methods and tools between software development and safety engineering today. We need to invest more in formality upfront - capturing safety work products in semantically rich models that are machine-processable, defining automatic consistency checks, and automating the generation of documentation - to benefit later. Transferring best practices from software to safety engineering is worth exploring. We advocate for safety factories, which integrate safety tooling and methods into software development pipelines.

cs.SE↗

arXiv2025

GlobDB: A comprehensive species-dereplicated microbial genome resource

Over the past years, substantial numbers of microbial species' genomes have been deposited outside of conventional INSDC databases. The GlobDB aggregates 14 independent genomic catalogues to provide a comprehensive database of species-dereplicated microbial genomes, with consistent taxonomy, annotations, and additional analysis resources. The GlobDB is available at https://globdb.org/.

q-bio.GN↗

arXiv2024

Fully printed flexible perovskite solar modules with improved energy alignment by tin oxide surface modification

Fully printed flexible perovskite solar cells (f-PSCs) show great potential for the commercialization of perovskite photovoltaics owing to their compatibility with high-throughput roll-to-roll (R2R) production. However, the challenge remains in the deficiency in controlling interfacial recombination losses of the functional layer, causing remarkable loss of power conversion efficiency (PCE) in industrial production. Here, a fullerene-substituted alkylphosphonic acid dipole layer is introduced between the R2R-printed tin oxide electron transport layer and the perovskite active layer to reduce the energetic barrier and to suppress surface recombination at the buried interface. The resulting f-PSCs exhibit a PCE of 17.0% with negligible hysteresis, retain 95% of their initial PCE over 3000 bending cycles and achieve a T95 lifetime of 1200 h under 1 sun and 65 degreeC in nitrogen atmosphere. Moreover, the fully printed flexible perovskite solar mini-modules (f-PSMs) with a 20.25 cm2 aperture area achieve a PCE of 11.6%. The encapsulated f-PSMs retain 90% of their initial PCE after 500 h damp-heat testing at 65 degreeC and 85% relative humidity (ISOS-D3). This work marks an important progress toward the realization of efficient and stable flexible perovskite photovoltaics for commercialization.

cond-mat.mtrl-sci↗

arXiv2024

The Open Autonomy Safety Case Framework

A system safety case is a compelling, comprehensible, and valid argument about the satisfaction of the safety goals of a given system operating in a given environment supported by convincing evidence. Since the publication of UL 4600 in 2020, safety cases have become a best practice for measuring, managing, and communicating the safety of autonomous vehicles (AVs). Although UL 4600 provides guidance on how to build the safety case for an AV, the complexity of AVs and their operating environments, the novelty of the used technology, the need for complying with various regulations and technical standards, and for addressing cybersecurity concerns and ethical considerations make the development of safety cases for AVs challenging. To this end, safety case frameworks have been proposed that bring strategies, argument templates, and other guidance together to support the development of a safety case. This paper introduces the Open Autonomy Safety Case Framework, developed over years of work with the autonomous vehicle industry, as a roadmap for how AVs can be deployed safely and responsibly.

cs.SE↗

arXiv2024

The role of fluid friction in streamer formation and biofilm growth

Bacillus subtilis biofilms were grown in laminar channel flow at wall shear stress spanning one order of magnitude ($\tau_w = 0.068$ Pa to $\tau_w = 0.67$ Pa). We monitor, non-invasively, the evolution of the three-dimensional distribution of biofilm over seven days using optical coherence tomography (OCT). The obtained biofilms consist of many microcolonies where the characteristic colony has a base structure in the form of a leaning pillar and a streamer in the form of a thin filament that originates near the tip of the pillar. While the shape, size and distribution of these microcolonies depend on the imposed shear stress, the same structural features appear consistently for all shear stress values. The formation of streamers seems to occur after the development of a base structure, suggesting that the latter induces a curved secondary flow that triggers the formation of the streamers. Moreover, we observe that the biofilm volume grows approximately linearly over seven days for all the shear stress values, with a growth rate that is inversely proportional to the wall shear stress. We develop a simple model of friction-limited growth, which agrees with the experimental observations. The model provides physical insight into growth mechanisms and can be used to develop accurate continuum models of bacterial biofilm growth.

cond-mat.soft↗

arXiv2023

Modeling the Ratio of Correlated Biomarkers Using Copula Regression

Modeling the ratio of two dependent components as a function of covariates is a frequently pursued objective in observational research. Despite the high relevance of this topic in medical studies, where biomarker ratios are often used as surrogate endpoints for specific diseases, existing models are based on oversimplified assumptions, assuming e.g.\@ independence or strictly positive associations between the components. In this paper, we close this gap in the literature and propose a regression model where the marginal distributions of the two components are linked by Frank copula. A key feature of our model is that it allows for both positive and negative correlations between the components, with one of the model parameters being directly interpretable in terms of Kendall's rank correlation coefficient. We study our method theoretically, evaluate finite sample properties in a simulation study and demonstrate its efficacy in an application to diagnosis of Alzheimer's disease via ratios of amyloid-beta and total tau protein biomarkers.

stat.ME↗

arXiv2023

Guidelines for Material Design in Semitransparent Organic Solar Cells

Organic solar cells (OSCs) are uniquely suited for semitransparent applications due to their adjustable absorption spectrum. However, most high-performance semitransparent cells reported to date are based on materials that have shown high power conversion efficiency for opaque devices. We therefore present a model to assess the optimum efficiency and transparency for a specific donor and acceptor band gap. The absorption characteristics of both donor and acceptor are modeled with spectral data of typical absorber materials from the literature which are adjusted to achieve the desired band gap value. The results show three distinct regions of high light utilization efficiency if the photopic curve is employed as a weighting function (corresponding to window applications), and a broad maximum for the plant action spectrum as a weighting function (corresponding to greenhouse applications). When comparing these findings to reported experimental values, it is evident that the band gaps of the materials used for the experimental studies do not correspond to the maxima identified by our simulation model. The analysis of the energy levels of molecules recorded in the literature confirms that all band gaps and therefore all LUE maxima are chemically feasible so that the performance of semitransparent OSCs can be further improved by designing materials with optimized absorption spectra.

cond-mat.mtrl-sci↗

arXiv2023

Matching the photocurrent of perovskite/organic tandem solar modules by varying the cell width

Photocurrent matching in conventional monolithic tandem solar cells is achieved by choosing semiconductors with complementary absorption spectra and by carefully adjusting the optical properties of the complete top and bottom stacks. However, for thin film photovoltaic technologies at the module level, another design variable significantly alleviates the task of photocurrent matching, namely the cell width, whose modification can be readily realized by the adjustment of the module layout. Herein we demonstrate this concept at the experimental level for the first time for a 2T-mechanically stacked perovskite (FAPbBr3)/organic (PM6:Y6:PCBM) tandem mini-module, an unprecedented approach for these emergent photovoltaic technologies fabricated in an independent manner. An excellent Isc matching is achieved by tuning the cell widths of the perovskite and organic modules to 7.22 mm (PCEPVKT-mod= 6.69%) and 3.19 mm (PCEOPV-mod= 12.46%), respectively, leading to a champion efficiency of 14.94% for the tandem module interconnected in series with an aperture area of 20.25 cm2. Rather than demonstrating high efficiencies at the level of small lab cells, our successful experimental proof-of-concept at the module level proves to be particularly useful to couple devices with non-complementary semiconductors, either in series or in parallel electrical connection, hence overcoming the limitations imposed by the monolithic structure.

cond-mat.mtrl-sci↗

arXiv2023

Cutting 'lab-to fab' short: High Throughput Optimization and Process Assessment in Roll-to-Roll Slot Die Coating of Printed Photovoltaics

Commercialization of printed photovoltaics requires knowledge of the optimal composition and microstructure of the single layers, and the ability to control these properties over large areas under industrial conditions. While microstructure optimization can be readily achieved by lab scale methods, the transfer from laboratory scale to a pilot production line ('lab to fab') is a slow and cumbersome process. Here, we show how we can optimize the performance of organic solar cells and at the same time assess process performance in a 2D combinatorial approach directly on an industrially relevant slot die coating line. This is enabled by a multi-nozzle slot die coating head allowing parameter variations along and across the web. This modification allows us to generate and analyze 3750 devices in a single coating run, varying the active layer donor:acceptor ratio and the thickness of the electron transport layer (ETL). We use Gaussian Process Regression (GPR) to exploit the whole dataset for precise determination of the optimal parameter combination. Performance-relevant features of the active layer morphology are inferred from UV-Vis absorption spectra. By mapping morphology in this way, small undesired gradients of process conditions (extrusion rates, annealing temperatures) are detected and their effect on device performance is quantified. The correlation between process parameters, morphology and performance obtained by GPR provides hints to the underlying physics, which are finally quantified by automated high-throughput drift-diffusion simulations. This leads to the conclusion that voltage losses which are observed for very thin ETL coatings are due to incomplete coverage of the electrode by the ETL, which cause enhanced surface recombination.

physics.app-ph↗

arXiv2022

Mid-Infrared Photothermal-Fluorescence in Situ Hybridization for Functional Analysis and Genetic Identification of Single Cells

Simultaneous identification and metabolic analysis of microbes with single-cell resolution and high throughput is necessary to answer the question of "who eats what, when, and where" in complex microbial communities. Here, we present a mid-infrared photothermal-fluorescence in situ hybridization (MIP-FISH) platform that enables direct bridging of genotype and phenotype. Through multiple improvements of MIP imaging, the sensitive detection of isotopically-labelled compounds incorporated into proteins of individual bacterial cells became possible, while simultaneous detection of FISH labelling with rRNA-targeted probes enabled the identification of the analyzed cells. In proof-of-concept experiments, we showed that the clear spectral red shift in the protein amide I region due to incorporation of $^{13}$C atoms originating from $^{13}$C-labelled-glucose can be exploited by MIP-FISH to discriminate and identify $^{13}$C-labelled bacterial cells within a complex human gut microbiome sample. The presented methods open new opportunities for single-cell structure-function analyses for microbiology.

physics.bio-ph↗

arXiv2021

Characterizing Idioms: Conventionality and Contingency

Idioms are unlike most phrases in two important ways. First, the words in an idiom have non-canonical meanings. Second, the non-canonical meanings of words in an idiom are contingent on the presence of other words in the idiom. Linguistic theories differ on whether these properties depend on one another, as well as whether special theoretical machinery is needed to accommodate idioms. We define two measures that correspond to the properties above, and we implement them using BERT (Devlin et al., 2019) and XLNet(Yang et al., 2019). We show that idioms fall at the expected intersection of the two dimensions, but that the dimensions themselves are not correlated. Our results suggest that special machinery to handle idioms may not be warranted.

cs.CL↗

arXiv2021

ProsoBeast Prosody Annotation Tool

The labelling of speech corpora is a laborious and time-consuming process. The ProsoBeast Annotation Tool seeks to ease and accelerate this process by providing an interactive 2D representation of the prosodic landscape of the data, in which contours are distributed based on their similarity. This interactive map allows the user to inspect and label the utterances. The tool integrates several state-of-the-art methods for dimensionality reduction and feature embedding, including variational autoencoders. The user can use these to find a good representation for their data. In addition, as most of these methods are stochastic, each can be used to generate an unlimited number of different prosodic maps. The web app then allows the user to seamlessly switch between these alternative representations in the annotation process. Experiments with a sample prosodically rich dataset have shown that the tool manages to find good representations of varied data and is helpful both for annotation and label correction. The tool is released as free software for use by the community.

eess.AS↗

arXiv2020

Improving 3D convolutional neural network comprehensibility via interactive visualization of relevance maps: Evaluation in Alzheimer's disease

Background: Although convolutional neural networks (CNN) achieve high diagnostic accuracy for detecting Alzheimer's disease (AD) dementia based on magnetic resonance imaging (MRI) scans, they are not yet applied in clinical routine. One important reason for this is a lack of model comprehensibility. Recently developed visualization methods for deriving CNN relevance maps may help to fill this gap. We investigated whether models with higher accuracy also rely more on discriminative brain regions predefined by prior knowledge. Methods: We trained a CNN for the detection of AD in N=663 T1-weighted MRI scans of patients with dementia and amnestic mild cognitive impairment (MCI) and verified the accuracy of the models via cross-validation and in three independent samples including N=1655 cases. We evaluated the association of relevance scores and hippocampus volume to validate the clinical utility of this approach. To improve model comprehensibility, we implemented an interactive visualization of 3D CNN relevance maps. Results: Across three independent datasets, group separation showed high accuracy for AD dementia vs. controls (AUC$\geq$0.92) and moderate accuracy for MCI vs. controls (AUC$\approx$0.75). Relevance maps indicated that hippocampal atrophy was considered as the most informative factor for AD detection, with additional contributions from atrophy in other cortical and subcortical regions. Relevance scores within the hippocampus were highly correlated with hippocampal volumes (Pearson's r$\approx$-0.86, p<0.001). Conclusion: The relevance maps highlighted atrophy in regions that we had hypothesized a priori. This strengthens the comprehensibility of the CNN models, which were trained in a purely data-driven manner based on the scans and diagnosis labels.

eess.IV↗