arXiv subjects

Justin Liu

Publications and source records attributed to Justin Liu.

6 recordsLinked to original sources

arXiv2026

Beyond Generalist LLMs: Specialist Agentic Systems for Structured Code Workflow Execution

Large Language Models (LLMs) have accelerated the adoption of software development agents, now widely available as Integrated Development Environment (IDE) extensions and standalone applications. While these agents are typically general-purpose, it remains unclear whether specialist agents justify their additional development effort. We investigate this question in the context of business process automation, focusing on the transformation of Business Process Model and Notation (BPMN) diagrams into executable agentic workflows. Since BPMN specifies explicit control-flow semantics, we focus on deterministic workflows in which a fixed process model and inputs uniquely determine the executed path. We introduce a specialist workflow for this task and compare it against generalist agents such as Roo and Cline. Our results show that the specialist solution produces agents that outperform generalist baselines by approximately 9-20 percentage points in tool-use exactness, 2-4x in penalty-adjusted latency, and 3x fewer tool-call errors, while reducing generation token cost by over 95% and eliminating repair iterations. We also find that generalist agents generate code inconsistently in both functionality and quality, limiting their suitability for industrial settings where reliability and maintainability are essential.

cs.SE↗

arXiv2025

Extendability of $1$-decomposable complexes

A well-known conjecture of Simon (1994) states that any pure $d$-dimensional shellable complex on $n$ vertices can be extended to $\Delta_{n-1}^{(d)}$, the $d$-skeleton of the $(n-1)$-dimensional simplex, by attaching one facet at a time while maintaining shellability. The notion of $k$-decomposability for simplicial complexes, which generalizes shellability, was introduced by Provan and Billera (1980). Coleman, Dochtermann, Geist, and Oh (2022) showed that any pure $d$-dimensional $0$-decomposable complex on $n$ vertices can similarly be extended to $\Delta_{n-1}^{(d)}$, attaching one facet at a time while preserving $0$-decomposability. In this paper, we investigate the analogous question for $1$-decomposable complexes. We prove a slightly relaxed version: any pure $d$-dimensional $1$-decomposable complex on $n$ vertices can be extended to $\Delta_{n + d - 3}^{(d)}$, attaching one facet at a time while maintaining $1$-decomposability.

math.CO↗

arXiv2025

CapyMOA: Efficient Machine Learning for Data Streams and Online Continual Learning in Python

CapyMOA is an open-source Python library for efficient machine learning on data streams and online continual learning. It provides a structured framework for real-time learning, supporting adaptive models that evolve over time. CapyMOA's architecture allows integration with frameworks such as MOA, scikit-learn and PyTorch, enabling the combination of high-performance online algorithms with modern deep learning techniques. By emphasizing efficiency, scalability, and usability, CapyMOA allows researchers and practitioners to tackle dynamic learning challenges across various domains. Website: https://capymoa.org. GitHub: https://github.com/adaptive-machine-learning/CapyMOA.

cs.LG↗

arXiv2023

On absolutely friendly measures on $\mathbb{Q}_S^d$

In this paper, we extend the work of Pollington and Velani in \cite{PV} to an $S$-arithmetic set-up, where $S$ is a finite set of valuations of $\mathbb{Q}$. In particular, for an `absolutely friendly' measure supported on a compact set in $\mathbb{Q}_S^d$, we give a summation condition on approximating function $\psi$ such that $\mu$ almost no point in the compact set is $\psi$ approximable. The crucial ingredient is a version of the simplex lemma that we prove dynamically.

math.NT↗

arXiv2022

A Novel Implementation of Machine Learning for the Efficient, Explainable Diagnosis of COVID-19 from Chest CT

In a worldwide health crisis as exigent as COVID-19, there has become a pressing need for rapid, reliable diagnostics. Currently, popular testing methods such as reverse transcription polymerase chain reaction (RT-PCR) can have high false negative rates. Consequently, COVID-19 patients are not accurately identified nor treated quickly enough to prevent transmission of the virus. However, the recent rise of medical CT data has presented promising avenues, since CT manifestations contain key characteristics indicative of COVID-19. This study aimed to take a novel approach in the machine learning-based detection of COVID-19 from chest CT scans. First, the dataset utilized in this study was derived from three major sources, comprising a total of 17,698 chest CT slices across 923 patient cases. Image preprocessing algorithms were then developed to reduce noise by excluding irrelevant features. Transfer learning was also implemented with the EfficientNetB7 pre-trained model to provide a backbone architecture and save computational resources. Lastly, several explainability techniques were leveraged to qualitatively validate model performance by localizing infected regions and highlighting fine-grained pixel details. The proposed model attained an overall accuracy of 0.927 and a sensitivity of 0.958. Explainability measures showed that the model correctly distinguished between relevant, critical features pertaining to COVID-19 chest CT images and normal controls. Deep learning frameworks provide efficient, human-interpretable COVID-19 diagnostics that could complement radiologist decisions or serve as an alternative screening tool. Future endeavors may provide insight into infection severity, patient risk stratification, and prognosis.

eess.IV↗

arXiv2007

Nonclassicality of quantum excitation of classical coherent fields in thermal environments

The nonclassicality of photon-added coherent fields in the thermal channel is investigated by exploring the volume of the negative part of the Wigner function which reduces with the dissipative time. The Wigner functions become positive when the decay time exceeds a threshold value. For the case of the single photon-added coherent state, we derive the exact threshold values of decay time in the thermal channel. For arbitrary partial negative Wigner distribution function, a generic analytical relation between the mean photon number of heat bath and the threshold value of decay time is presented. Finally, the possible application of SPACSs in quantum computation has been briefly discussed. OCIS codes: 270.0270, 270.2500, 000.5490

quant-ph↗