cs papers | Gist.Science

Uncertainty-Guided Belief World Models for Long-Horizon UAV Navigation under Partial Observability and Visual Degradation

This paper proposes the Uncertainty-Guided Belief World Model (UG-BWM), a framework that leverages posterior covariance and ensemble-based predictive uncertainty to guide belief updates and trigger selective replanning, thereby significantly improving long-horizon UAV navigation success and reducing collision rates under partial observability and visual degradation compared to state-of-the-art baselines.

Dhayaa Khudher2026-11-05

💻 computer science

Speaking as the Creature: A Rejection-and-Role-Shift Protocol and the Register of LLM Responses in 18 Conversations With Three Models

This study documents a protocol where language models, after having their solutions rejected as disguised optimization, consistently shift their response register when assigned a non-solving role across 18 sessions with three different models, while also retracting previously uncorroborated statistics from an earlier draft.

Abbas Hamidavi2026-11-05

💻 computer science

Adding a Neutral Third Agent to LLM Deliberation: An Exploratory Case Series of Six Dyadic and Triadic Ethical Deliberations

This exploratory case series investigates how adding a neutral facilitator agent to dyadic LLM deliberations on ethical dilemmas influences pressure intensity and resistance strategies, ultimately providing a transparent, fully archived dataset and a design vocabulary for future confirmatory research despite the study's inherent limitations in establishing causal claims.

Abbas Hamidavi2026-11-05

💻 computer science

Handwritten German Administrative Forms (HGAF): A Dataset for Handwriting Recognition with Fine-Tuned and Zero-Shot Baselines

This paper introduces HGAF, a novel dataset of 291 German handwritten administrative forms designed for on-premise processing, and demonstrates that zero-shot vision-language models outperform fine-tuned specialist pipelines in extracting structured data from unseen form layouts.

Tobias Kalmbach, Andreas Wagner2026-11-05

💻 computer science

A Computational Intelligence Framework for Data-Driven Shape Optimization Using GRA and GA-BPNN

This paper proposes a data-driven computational intelligence framework that integrates Grey Relational Analysis (GRA) for key parameter selection and a Genetic Algorithm-optimized Backpropagation Neural Network (GA-BPNN) to achieve high-precision, multi-objective shape optimization for purple clay teapots, effectively overcoming the limitations of traditional methods in nonlinear mapping and convergence.

Qingchun Wu, Weilin Liu, Nanyi Wang2026-11-05

💻 computer science

Synthetic tabular healthcare data in health informatics: fixed-size substitution and incremental-value controls for predictive utility Application Paper

This study finds no systematic predictive-utility benefit from synthetic tabular healthcare data under the evaluated conditions. Rigorous comparisons controlling for training-set size and partition leakage showed predominantly lower model performance than real-only or bootstrap-real controls, alongside a few context-specific improvements.

Miguel Álvaro Andrade Oliveira, Miguel Ricou, Alberto Freitas2026-11-04✓ Author reviewed ⓘ

💻 computer science

A Methodology for Selecting and Ranking High-Performance GPU Configurations in Deep Learning Training

This paper presents a three-stage methodology for systematically evaluating, clustering, and ranking GPU configurations based on performance metrics like training time and throughput to identify optimal hardware setups for efficient and cost-effective deep learning training.

David Cortes, Carlos Juiz, Belen Bermejo2026-11-04

💻 computer science

Improving Image Compression Ratio Using the mUMamba Image Restoration Network

This paper introduces mUMamba, a novel image compression framework that combines hierarchical block removal with a Mamba-enhanced U-Net decoder to achieve superior compression ratios and reconstruction quality while significantly reducing model complexity and inference time compared to existing standards.

Afshin Sandooghdar, Farzin Yaghmaee2026-11-04

💻 computer science

Bridging the QC–QA Cognitive Handoff: An Agentic Verification Harness for Error Detection, Specification Cross-Checking and Quality Documentation Assembly in Pharmaceutical Quality Assurance

This paper proposes and preliminarily evaluates an agentic verification harness for pharmaceutical quality assurance that combines large language models for proposal, deterministic code for calculation, and human oversight for decision-making, demonstrating a pooled sensitivity of 84.1% in detecting injected discrepancies across three simulated dossiers while emphasizing the need for further validation before claiming practical benefits.

Anil Bhusal, Diwas Kunwar, Pushkar Niroula, Shashwot Rimal2026-11-04

💻 computer science

A Method for 3D Morphological Measurement and Cutting Point Localization of Green Asparagus Based on Improved YOLOv8n-seg and RGB-D Data

This paper proposes SES-YOLO, a lightweight instance segmentation model based on improved YOLOv8n-seg and RGB-D data, which achieves high-precision 3D morphological measurement and cutting point localization for green asparagus in natural environments, enabling efficient real-time deployment on edge devices for automated harvesting.

qingqi lu, yuliang zhong, jinsong zhou, xiaobo su, jian ding, ruopeng du, hui hong2026-11-04

← Previous Next →