💻 Category

cs

4158 papers

💻 computer science

Leakage-audited cross-dataset evaluation reveals large generalization gaps in deep learning models for apple and maize disease recognition

This study demonstrates that deep learning models for apple and maize disease recognition, despite achieving near-perfect accuracy on curated benchmarks after rigorous leakage auditing, suffer from severe generalization gaps when tested on independently collected external datasets, highlighting the critical need for independent evaluation to assess true field readiness.

Yogesh ., P Sudir, Princy Randhawa, Muzameel Ahmed, B M Chandrakala2026-11-06
💻 computer science

Simulated Users Can Reproduce an Algorithm Audit Without Sharing the Mechanism That Produced the Finding

This study proposes and validates a novel procedure for algorithm audits using simulated users that does not require real user data, demonstrating that while large language models can reproduce specific audit findings, their validity depends critically on the number of simulation runs and the method of preference conveyance, necessitating stricter pre-expenditure tests and stability checks.

Keito Inoshita2026-11-06
💻 computer science

One Successful Run Is Not an Exploit: Replaying LLM-Written Exploit Scripts with the Model Out of the Loop

This paper demonstrates that single-run success in LLM-based offensive security benchmarks is often misleading by showing that while generated exploit scripts are highly stable when replayed without the model, the actual instability lies in the generation process rather than the scripts themselves, necessitating a separation of model performance from artifact reliability for accurate evaluation.

Wensheng Huang2026-11-06
💻 computer science

Network-Centric Ransomware Intrusion Detection Framework

This paper proposes a network-centric ransomware detection framework utilizing a novel "Traffic of Interest" (ToI) analysis technique and machine learning to achieve near-perfect precision and zero false negatives in identifying victim nodes, while also providing a repository of unique packet features for detecting specific ransomware families that evade host-based tools.

Gervais Hatungimana, Abdun Naser Mahmood, Mohammad Jabed Morshed Chowdhury2026-11-06
💻 computer science

Multimodal Prompt Learning with Vision--Language Foundation Models for Automated Storyboard Assessment and Formative Feedback in Film Education

This paper introduces CRaFT, a multimodal prompt learning framework that leverages frozen vision–language foundation models and criterion-constrained optimal transport to automate storyboard assessment and generate evidence-based formative feedback, achieving state-of-the-art performance across multiple public datasets while supporting teacher-aligned evaluation protocols.

Lu Zhang2026-11-06
💻 computer science

3D Ischemic Stroke Lesion Segmentation in MRI with Attention U-Net and Voxel-Level Uncertainty Quantification

This paper proposes a reliability-aware 3D framework that combines an Attention U-Net with Monte Carlo Dropout to perform ischemic stroke lesion segmentation on MRI while simultaneously generating voxel-level uncertainty maps that correlate with prediction errors and improve recall without additional training.

Outmane Labaybi, Mohamed Bennani Taj, Khalid Elfahssi, Rachid Flouchi, Mohamed EL FAR2026-11-06
💻 computer science

Network Based Quantum Inspired Deep Learning Framework for Cardiovascular Disease Prediction Using Interpretable Feature Interaction Networks

This paper proposes a hybrid quantum-inspired deep learning framework that utilizes ZZ-feature maps to model clinical feature interactions for cardiovascular disease prediction, achieving 90.5% accuracy and high interpretability on the UCI Cleveland dataset while acknowledging the need for future ablation studies to isolate specific component contributions.

Ritu Singh, Shaheen Naz, Himanshu Payal, Dinesh Kumar Nishad2026-11-06