arXiv subjects

Niruthiha Selvanayagam

Publications and source records attributed to Niruthiha Selvanayagam.

6 recordsLinked to original sources

arXiv2026

AI-to-AI Code Reviews of GitHub Pull Requests

AI coding agents are increasingly integrated into software development workflows, operating on both sides of the pull-request (PR) process: AI authoring agents create or modify PRs, while AI reviewers evaluate them. This creates a closed loop in which one AI coding agent reviews a contribution attributed to another. We construct a large-scale dataset of AI-to-AI code review by linking AI-attributed PRs with AI-attributed review events from CodAGE, a public dataset of coding-agent-generated GitHub events. Our dataset contains 248,641 unique AI-attributed PRs that received at least one AI-attributed review. Of these, 45,269 received cross-product review and 208,145 received same-product review; 4,773 PRs received both. Cross-product AI-to-AI review occurred in approximately 1.6% of identified agent-authored PRs but was substantial in absolute terms, and its volume increased by more than two orders of magnitude from 2025-Q1 to 2025-Q3. Reviewer output varied across author-reviewer configurations. CodeRabbit labeled 35.0% of its comments on Claude Code-authored PRs as refactor comments, compared with 10.5% on Copilot-authored PRs, although this difference may reflect characteristics of the PRs rather than the reviewer. For three of four dual-role reviewers, mean comments per PR were 58-65% higher in the same-product group, although effect sizes were small or negligible and the difference was concentrated in the upper tail. Among pairs with complete, nonnegative timestamps, the observed median latency was 1.2 minutes for cross-product pairs and 4.7 minutes for same-product pairs; differential timestamp availability and reviewer composition limit this comparison. Overall, closed-loop AI-to-AI review is increasing but remains a minority of identified agent activity, with review output varying across authoring-agent groups and product configurations.

cs.SE↗

arXiv2026

Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection

As Large Multimodal Models (LMMs) become integral to daily digital life, understanding their safety architectures is a critical problem for AI Alignment. This paper presents a systematic analysis of OpenAI's GPT-4o mini, a globally deployed model, on the difficult task of multimodal hate speech detection. Using the Hateful Memes Challenge dataset, we conduct a multi-phase investigation on 500 samples to probe the model's reasoning and failure modes. Our central finding is the experimental identification of a "Unimodal Bottleneck," an architectural flaw where the model's advanced multimodal reasoning is systematically preempted by context-blind safety filters. A quantitative validation of 144 content policy refusals reveals that these overrides are triggered in equal measure by unimodal visual 50% and textual 50% content. We further demonstrate that this safety system is brittle, blocking not only high-risk imagery but also benign, common meme formats, leading to predictable false positives. These findings expose a fundamental tension between capability and safety in state-of-the-art LMMs, highlighting the need for more integrated, context-aware alignment strategies to ensure AI systems can be deployed both safely and effectively.

cs.LG↗

arXiv2026

FragBench: Cross-Session Attacks Hidden in Benign-Looking Fragments

An attacker can split a malicious goal into sub-prompts that each look benign on their own and only become harmful in combination. Existing LLM safety benchmarks evaluate prompts one at a time, or across turns of a single chat, and so do not look for a malicious signal spread across separate sessions with no shared context. We build FragBench, a benchmark drawn from 24 real-world cyber-incident campaigns, which keeps the full attack trail: the multi-fragment kill chain, the per-fragment safety-judge verdicts, sandboxed execution traces, and a matched set of benign cover sessions. FragBench splits this trail into two paired tasks: an adversarial rewriter that hardens fragments against a single-turn safety judge (FragBench Attack), and a graph-based user-level detector trained on the resulting interactions (FragBench Defense). The single-turn judge is near chance on the released corpus by construction, but four GNN variants and three classical-ML baselines all recover the cross-session feature, reaching aggregate event-level F1 = 0.88-0.96. Defending against fragmented LLM misuse therefore requires modeling the cross-session interaction graph, rather than isolated prompts. Our generator, rewriter, sandbox harness, and detector are released at https://github.com/LidaSafety/fragbench.

cs.CR↗

arXiv2026

MLmisFinder: A Specification and Detection Approach of Machine Learning Service Misuses

Machine Learning (ML) cloud services, offered by leading providers such as Amazon, Google, and Microsoft, enable the integration of ML components into software systems without building models from scratch. However, the rapid adoption of ML services, coupled with the growing complexity of business requirements, has led to widespread misuses, compromising the quality, maintainability, and evolution of ML service-based systems. Though prior research has studied patterns and antipatterns in service-based and ML-based systems separately, automatic detection of ML service misuses remains a challenge. In this paper, we propose MLmisFinder, an automatic approach to detect ML service misuses in software systems, aiming to identify instances of improper use of ML services to help developers properly integrate ML components in ML service-based systems. We propose a metamodel that captures the data needed to detect misuses in ML service-based systems and apply a set of rule-based detection algorithms for seven misuse types. We evaluated MLmisFinder on 107 software systems collected from open-source GitHub repositories and compared it with a state-of-the-art baseline. Our results show that MLmisFinder effectively detects ML service misuses, achieving an average precision of 96.7\% and recall of 97\%, outperforming the state-of-the-art baseline. MLmisFinder also scaled efficiently to detect misuses across 817 ML service-based systems and revealed that such misuses are widespread, especially in areas such as data drift monitoring and schema validation.

cs.SE↗

arXiv2026

Self-Admitted Technical Debt in LLM Software: An Empirical Comparison with ML and Non-ML Software

Self-admitted technical debt (SATD), referring to comments flagged by developers that explicitly acknowledge suboptimal code or incomplete functionality, has received extensive attention in machine learning (ML) and traditional (Non-ML) software. However, little is known about how SATD manifests and evolves in contemporary Large Language Model (LLM)-based systems, whose architectures, workflows, and dependencies differ fundamentally from both traditional and pre-LLM ML software. In this paper, we conduct the first empirical study of SATD in the LLM era, replicating and extending prior work on ML technical debt to modern LLM-based systems. We compare SATD prevalence across LLM, ML, and non-ML repositories across a total of 477 repositories (159 per category). We perform survival analysis of SATD introduction and removal to understand the dynamics of technical debt across different development paradigms. Surprisingly, despite their architectural complexity, our results reveal that LLM repositories accumulate SATD at similar rates to ML systems (3.95% vs. 4.10%). However, we observe that LLM repositories remain debt-free 2.4x longer than ML repositories (a median of 492 days vs. 204 days), and then start to accumulate technical debt rapidly. Moreover, our qualitative analysis of 377 SATD instances reveals three new forms of technical debt unique to LLM-based development that have not been reported in prior research: Model-Stack Workaround Debt, Model Dependency Debt, and Performance Optimization Debt. Finally, by mapping SATD to stages of the LLM development pipeline, we observe that debt concentrates

cs.SE↗

arXiv2025

Multidimensional Analysis of Specific Language Impairment Using Unsupervised Learning Through PCA and Clustering

Specific Language Impairment (SLI) affects approximately 7 percent of children, presenting as isolated language deficits despite normal cognitive abilities, sensory systems, and supportive environments. Traditional diagnostic approaches often rely on standardized assessments, which may overlook subtle developmental patterns. This study aims to identify natural language development trajectories in children with and without SLI using unsupervised machine learning techniques, providing insights for early identification and targeted interventions. Narrative samples from 1,163 children aged 4-16 years across three corpora (Conti-Ramsden 4, ENNI, and Gillam) were analyzed using Principal Component Analysis (PCA) and clustering. A total of 64 linguistic features were evaluated to uncover developmental trajectories and distinguish linguistic profiles. Two primary clusters emerged: (1) high language production with low SLI prevalence, and (2) limited production but higher syntactic complexity with higher SLI prevalence. Additionally, boundary cases exhibited intermediate traits, supporting a continuum model of language abilities. Findings suggest SLI manifests primarily through reduced production capacity rather than syntactic complexity deficits. The results challenge categorical diagnostic frameworks and highlight the potential of unsupervised learning techniques for refining diagnostic criteria and intervention strategies.

cs.CL↗