cs papers | Gist.Science

Political Stance Detection on X During the 2024U.S. Election: A Comparative Study of Classical, Neural, and Transformer Models

This paper presents a comparative study of classical, neural, and transformer models for political stance detection on X during the 2024 U.S. election, demonstrating that while fine-tuned BERTweet achieves the highest accuracy, carefully engineered classical models like linear SVMs offer a highly competitive, more efficient, and interpretable alternative for scalable deployment.

Qile Wang, Sahar Ostadrahimi, Safoura Faghri, Mohammad Baksh, Matthew Louis Mauriello, Kenneth E. Barner2026-07-17

💻 computer science

Classical Methods Match or Exceed Two Recent Graph Neural Networks for Bipartite Community Detection Using Network Topology Alone

This paper demonstrates that classical community detection methods consistently match or outperform recent Graph Neural Networks on bipartite networks using only topology, based on a comprehensive evaluation of fourteen methods across eight real-world and five synthetic datasets.

Aneesh K Sajan2026-07-16

💻 computer science

A Workflow for Grant Discovery and Proposal Development Using Large Language Models: Development and Formative Evaluation

This paper describes and formative-evaluates a governed, large language model-supported workflow implemented at a Colombian health-AI company to systematically improve the throughput and traceability of grant discovery and proposal development while maintaining researcher oversight. In Phase 1 ("The Treasure Hunt"), the 83 candidates are identified as post-filter selections that passed the filtering criteria and were considered suitable, rather than being rejected.

Katherine Monsalve Barrientos, Natalia Castano-Villegas, Jose Zea, Laura Velásquez2026-07-16

💻 computer science

Iterative Feedback--Refinement for Faithful Structured Clinical-Note Representation

This paper proposes an iterative feedback-refinement framework using large language models to dynamically induce adaptive schemas for converting unstructured clinical notes into faithful, structured representations, which outperforms existing baselines by significantly reducing omission and hallucination without requiring labeled data or fine-tuning.

Eslam Ahmed Mohamed, Irini Logothetis, Adrian Bingham, Kon Mouzakis2026-07-16

💻 computer science

Hidden Universal Collapse Behind Apparent Robustness: Dual-Metric Cross-Site Audit of a Laparoscopic Surgical AI Detector

This study demonstrates that single-metric evaluations can mask the universal collapse of surgical AI models across sites by highlighting apparent robustness in specific classes, whereas dual-metric audits reveal widespread failure and that backbone substitution, in the configurations we tested, did not close the gap.

hui zhu, congbin zhu, Sio Lam UN, QI CHENG, ZHANDONG MENG, qiliang WANG, cong hu, huiying Zhu2026-07-16

💻 computer science

Automated Arrhythmia Detection from ECG Signals: AComparative Study of Bidirectional Long Short-Term MemoryNetworks and Residual Convolutional Neural Networks

This study benchmarks Bidirectional Long Short-Term Memory networks with attention against Residual Convolutional Neural Networks for automated ECG arrhythmia detection, finding that the ResNet architecture achieves superior performance (98.0% accuracy) by better extracting complex morphological features from a dataset of over 109,000 heartbeats.

Jajang Jaya Purnama, Sri Rahayu, Risnandar Risnandar2026-07-16

💻 computer science

guardrail-rs A Fail-Open Reverse Proxy for Prompt-Injection Defense and PII Redaction in LLM Applications

This paper introduces guardrail-rs, an open-source Rust-based reverse proxy designed to protect LLM applications by composable detection of prompt injections and PII redaction with a fail-open architecture, while transparently reporting its performance characteristics and real-world engineering challenges.

Min Htet Myet2026-07-16

💻 computer science

A Deterministic Spatial Representation Packet for Sports Trajectories Under Explicit Coordinate, Threshold, and Tolerance Assumptions

This paper presents a deterministic pipeline for converting finite, noisy sports trajectory samples into auditable symbolic artifacts by explicitly enforcing coordinate, threshold, and tolerance assumptions, while strictly limiting its scope to representation and traceability rather than physical reconstruction or semantic inference.

WU YANG CHEN XI2026-07-16

💻 computer science

The Use of Artificial Intelligence in Article Searching and Scientific Writing: Evidence, Impact, and Future Directions

This review synthesizes evidence showing that while AI tools significantly enhance efficiency and clarity in literature searching and scientific writing, their responsible adoption requires hybrid human-AI approaches to address persistent limitations in reliability, bias, and academic integrity.

Rafael Baptista, Guilherme Donadel2026-07-16

💻 computer science

Beyond Accuracy: A Comprehensive Evaluation of ResNet50 and Vision Transformer for Trustworthy Pneumonia Detection from Chest X-Ray Images

This study presents a comprehensive evaluation of ResNet50 and Vision Transformer (ViT-B16) for pneumonia detection from chest X-rays, revealing that while ResNet50 offers superior accuracy, calibration, and efficiency, ViT-B16 demonstrates greater robustness against image degradations, thereby highlighting the necessity of multi-dimensional assessment for trustworthy diagnostic systems.

VISHAL MANTA2026-07-16

← Previous Next →