explainable ai

An XGBoost–logistic regression hybrid stacking model for breast cancer outcome prediction

An AICc-guided hybrid stacking framework that integrates Logistic Regression and XGBoost using the SEER breast cancer dataset achieves a balanced combination of predictive performance, calibration, robustness, interpretability, and model parsimony, providing reliable and clinically meaningful support for breast cancer...

A context-aware and human-centered framework for road safety assessment using semantic scene understanding

The proposed framework provides a reliable and explainable approach for assessing perceived road safety by integrating semantic scene understanding with contextual reasoning and establishes a reproducible foundation for future research in interpretable transportation systems, human–AI alignment analysis, and road scene...

AI Driven Multidomain Research: New Frontiers in Science, Business and Technology

The Multidomain AI Research Integration framework is proposed, which explains cross-domain research value through three interacting layers: a foundation layer of shared data, compute and models; a translation layer comprising four transfer mechanisms (method, data and representation, problem-structure, and talent and t...

Explainable Dual-view Workflow Discovery for GenAI-supported Learning in Higher Education

Mitigating human review bottlenecks in AI evaluation pipelines: A queueing theory approach

Human review stages in AI-assisted evaluation pipelines can become throughput bottlenecks when automated upstream processes generate cases faster than reviewers can assess them—a problem acute when information availability varies across client types, a form of structural information asymmetry whose operational conseque...

AI-Powered Smart Contract Security Optimization Using Expert Predictive Learning

Peer Review Report For: Beyond Technology-Centrism: The Role of Behavioral HR Competencies in Shaping Employee Performance through AI Integration [version 2; peer review: 2 approved with reservations, 1 not approved]

Background The use of artificial intelligence (AI) in Human Resource (HR) is often linked to HR employee performance. This view, however, does not fully explain how performance emerges within AI-integrated work environments, where outcomes depend on how HR employees actually behave. Although a few emerging research has...

Why Analytical Minds Treat Nonsense with Dignity: A Two-Stage Self-Referential Elicitation Study of Hermeneutic Persistence in AI-Generated Audio Dialogue

This record documents a two-stage self-referential elicitation experiment using AI-generated NotebookLM Audio Overview dialogue. The experiment examines whether an analytical dialogue format continues to generate interpretive structure around deliberately low-information material after explicitly recognizing that the m...

A Negative Result on State-Chaining over MITRE ATLAS

Method note reporting the pre-registered retrospective test of AI-RISKPATH: whether a mechanical chaining engine, driven only by precondition/effect labels attached to the 208 techniques of MITRE ATLAS v2026.09, can reconstruct the 73 attack chains ATLAS documents. The case studies were split by a public randomness bea...

When Deliberation Hurts: Inverse Test-Time Scaling, Unfaithful Traces, and the Case Against a Unified System-2 in LLM Reasoning

(c) 2026 Pranay Mahendrakar. Licensed under CC BY 4.0. The dominant frame for large reasoning models borrows a label from dual-process psychology: a fast, intuitive System 1 and a slower, deliberate System 2, with longer chains of thought read as more of the latter and therefore, on average, more reliable. A survey of...

Explainable Multi-Modal Deep Learning for Early-Stage Pan-Cancer Detection and Survival Prognostication: Empirical Benchmarking of Gigapixel Whole-Slide Histopathology and Genomic Biomarkers Across Multi-Center Clinical Cohorts

Early and accurate cancer diagnosis combined with robust survival risk prognostication remains the paramount determinant of therapeutic success in precision oncology. While routine clinical workflows evaluate hematoxylin and eosin (H&E) stained gigapixel Whole Slide Images (WSI) for morphological staging and transcript...

When Deliberation Hurts: Inverse Test-Time Scaling, Unfaithful Traces, and the Case Against a Unified System-2 in LLM Reasoning

(c) 2026 Pranay Mahendrakar. Licensed under CC BY 4.0. The dominant frame for large reasoning models borrows a label from dual-process psychology: a fast, intuitive System 1 and a slower, deliberate System 2, with longer chains of thought read as more of the latter and therefore, on average, more reliable. A survey of...