cs papers | Gist.Science

Evidence Constrained Agentic Retrieval Augmented Generation for Substation Civil Engineering Preliminary Design Documents in a Single Project Study

This paper proposes EC-ARAG, an evidence-constrained agentic retrieval-augmented generation framework that significantly enhances the reliability, consistency, and efficiency of substation civil engineering preliminary design drafting by integrating regulation-aware retrieval, conflict resolution, and dual verification loops to outperform conventional RAG systems.

Yizhang Huang, Xinhui Zhang2026-08-26

💻 computer science

Evaluation Pitfalls and Multimodal Baselines for the \dataset{} IoT Malware Dataset

This paper introduces the first systematic baselines and evaluation protocols for the CIC-YNU-IoTMal2026 multimodal IoT malware dataset, revealing critical pitfalls such as data leakage from random splits, high miss rates for dormant samples under leakage-free conditions, and severe cross-architecture fragility, while demonstrating that protocol choices rather than model selection ultimately dominate reported performance.

Xuetong Zhang, Yifei Xing, Zhibin Guo, Jianmin Li2026-08-26

💻 computer science

A Localization-Aware Heterogeneous CNN Ensemble with Neural Meta- Fusion for Plant Disease Classification, with a Component Analysis on Laboratory and Field Images

This paper presents a localization-aware heterogeneous CNN ensemble with neural meta-fusion that demonstrates how the critical components of plant disease classification pipelines shift from ensemble breadth and handcrafted descriptors in laboratory settings to leaf localization and learned fusion when applied to field imagery, validated through rigorous ablation studies on both PlantVillage and PlantDoc datasets.

Ehsan Gaballah2026-08-26

💻 computer science

Hybrid 3D-CNN, Bi-LSTM, and Transformer-Based Framework for Anomaly Detection of Human Behaviour in Video Surveillance with Adaptive Error Minimization

This paper proposes a novel hybrid deep learning framework that integrates 3D-CNN, Bi-LSTM, and Transformer architectures with an adaptive threshold module and multi-objective loss optimization to achieve state-of-the-art, real-time anomaly detection in video surveillance by significantly reducing false alarms across diverse environmental conditions.

Sameera yasam, Syed Ali Hussain, Vijaya Kumar Koppula2026-08-26

💻 computer science

CEGNN-Fraud: A Scalable Heterogeneous Temporal Graph Neural Network with Evidence-Constrained LLM Risk Reasoning

CEGNN-Fraud is a scalable heterogeneous temporal graph neural network that enhances financial fraud detection through an event-conditioned reliability gate and a constrained LLM for evidence-based risk reporting, demonstrating superior performance in accuracy, robustness, and generalization compared to existing baselines.

Yilin zhong2026-08-26

💻 computer science

A Framework for Dynamic Memory Management and Personalized Agent Decision-Making in Retrieval- Augmented Generation

This paper introduces DMAP-RAG, a framework that enhances Retrieval-Augmented Generation by integrating a tree-structured dual-storage memory architecture and personalized agent decision-making to achieve competitive performance in multi-turn dialogue coherence and user satisfaction.

Xiangguo Qian, Yong Xu, Chuqi Ma, Choon Kit Chan2026-08-26

💻 computer science

Brain tumor classification: A comparative analysis of deep learning techniques and impact of augmentation approaches on T1-CE MRI scans

This paper presents a comparative analysis of seven pre-trained deep learning models on T1-CE MRI scans for brain tumor classification, demonstrating that VGG-16 achieves the highest accuracy (92.45%) when trained with various augmentation techniques and optimizers, while utilizing Grad-CAM to enhance model interpretability.

Priyanka Sharma, Sakshi Ahuja2026-08-26

💻 computer science

A Self-Healing Architecture for Mitigating Bibliographic Hallucinations in LLM-Generated Academic Texts

This paper introduces a novel self-healing architecture that autonomously detects and mitigates bibliographic hallucinations in LLM-generated academic texts by combining multi-layered verification with an efficient sentence-based refinement process, achieving high identification accuracy (0.96) and significantly reducing computational costs while ensuring scholarly integrity.

Davide Tosi, Edoardo Crescenzo Mauriello, Andrea Utzeri2026-08-26

💻 computer science

When Evaluation Clarity Creates Algorithmic Anxiety: A Randomized Field Experiment of Explainable Fuzzy Assessment and Teacher Feedback Uptake

A randomized field experiment involving 982 Chinese university teachers demonstrates that explainable fuzzy AI feedback significantly enhances procedural fairness, calibrated trust, and subsequent instructional improvement compared to traditional or black-box AI systems, primarily by clarifying algorithmic uncertainty and fostering feedback uptake.

Yiran Zhou2026-08-26

💻 computer science

A Validated Measurement Protocol for Comparable, Cost-Aware Software Testing Evaluation: A Reproducible Benchmark, an Oracle Sampling-Budget Guarantee, and a Real-Fault Validity Study, Instantiated for Quantum Programs

This paper introduces a validated, reproducible measurement protocol (QSQ-Bench and Q-EVAL) that ensures comparable, cost-aware, and construct-valid software testing evaluations by establishing statistical oracle guarantees and demonstrating its effectiveness through a comprehensive study on quantum programs and a classical system.

Bhanwar Gupta, Sanjeev Rana2026-08-26

← Previous Next →