Institution profile

United Arab Emirates University

Academic institutionasia · ae
Official website
Research library51linked papers
Opportunities0open roles
Selected work

Representative Papers

α3-Bench: A Unified Benchmark of Safety, Robustness, and Efficiency for LLM-Based UAV Agents over 6G Networks

Jan 01, 2026arXiv.org

Current evaluation frameworks struggle to comprehensively assess the safety, protocol compliance, and task effectiveness of large language model (LLM)-driven drone agents under the dynamic constraints of 6G networks. To address this gap, this work proposes α³-Bench, a novel benchmark that integrates safety, robustness, and efficiency into a unified α³ evaluation metric. Built upon a multi-turn conversational control framework, α³-Bench features a dual-action-layer architecture enabling tool invocation and multi-agent collaboration, while incorporating 6G network emulation—including latency, jitter, and packet loss—and tool consistency verification. Evaluation across 17 prominent LLMs on a dataset of 113k dialogues reveals that although most models achieve high task success rates under nominal conditions, their robustness and communication efficiency degrade significantly under impaired 6G network conditions.

2 citationsRead paper

The 10th AI City Challenge

Aug 17, 2026

本文总结了第10届AI City挑战赛,该赛事通过结合基础模型、几何定位等方法解决智能交通和智慧城市中的多摄像头感知等问题。

0 citationsRead paper
Recent publications

Latest Papers

The 10th AI City Challenge

Aug 17, 2026

本文总结了第10届AI City挑战赛,该赛事通过结合基础模型、几何定位等方法解决智能交通和智慧城市中的多摄像头感知等问题。

0 citationsRead paper

ARISE: An adaptive residual-informed stability ensemble for feature selection in small-sample biomedical omics

Aug 14, 2026

This study addresses the challenge of balancing predictive performance, stability, and redundancy in feature selection for few-shot biomedical omics classification by proposing the ARISE framework. This method integrates multi-dimensional correlation signals with residual-aware redundancy control, employing adaptive weighted ensemble learning to jointly optimize multi-class discriminative capability and class-balanced stability. Across 210,000 evaluations, ARISE achieved top rankings across all metrics, attaining a balanced accuracy of 0.793 and a macro-F1 score of 0.776. These results demonstrate significant superiority over existing methods, effectively overcoming critical feature selection bottlenecks in few-shot molecular classification tasks.

0 citationsRead paper