8 papers
CoFiE: Coarse-to-Fine Evidence Selection for Efficient Streaming Video Understanding
Jing Jiang, Yiran Ling, Ruonan Li +2
Streaming video understanding requires Vision Language Models (VLLMs) to process growing video streams and answer user questions under tight latency constraints. Existing methods i…
H2Table: Hierarchical Hypergraph-Enhanced Large Language Models for Complex Table Reasoning
Jia Ling, Yangfan Wang, Chen Tang +4
Tables are ubiquitous across diverse domains, yet reasoning over them remains a significant challenge for modern large language models (LLMs). Current approaches typically lineariz…
Deploying and Evaluating a Smart-Agriculture Agentic Engine for Full-Season Soybean Farm Operations
Ao Qu, Panagiotis Michelakis, Linyuan Han +8
This paper presents FAIRY, a full-stack smart-agriculture agent system developed for and deployed to an operating soybean research farm at Harbin Institute of Technology's smart-ag…
CLASP: Closed-loop Asynchronous Spatial Perception for Open-vocabulary Desktop Object Grasping
Yiran Ling, Wenxuan Li, Siying Dong +5
Robot grasping of desktop object is widely used in intelligent manufacturing, logistics, and agriculture.Although vision-language models (VLMs) show strong potential for robotic ma…
HiEdit: Lifelong Model Editing with Hierarchical Reinforcement Learning
Yangfan Wang, Tianyang Sun, Chen Tang +3
Lifelong model editing (LME) aims to sequentially rectify outdated or inaccurate knowledge in deployed LLMs while minimizing side effects on unrelated inputs. However, existing app…
KA2L: A Knowledge-Aware Active Learning Framework for LLMs
Haoxuan Yin, Chen Tang, Yangfan Wang +2
Fine-tuning large language models (LLMs) with high-quality knowledge has been shown to enhance their performance effectively. However, there is a paucity of research on the depth o…
KCS: Diversify Multi-hop Question Generation with Knowledge Composition Sampling
Yangfan Wang, Jie Liu, Chen Tang +2
Multi-hop question answering faces substantial challenges due to data sparsity, which increases the likelihood of language models learning spurious patterns. To address this issue,…
GRAPHMOE: Amplifying Cognitive Depth of Mixture-of-Experts Network via Introducing Self-Rethinking Mechanism
Bo Lv, Chen Tang, Zifan Zheng +8
Traditional Mixture-of-Experts (MoE) networks benefit from utilizing multiple smaller expert models as opposed to a single large network. However, these experts typically operate i…