9 papers
GNFormer: Gaussian Neural Field Transformer for Efficient Camera-based Occupancy Prediction
Guangcheng Chen, Li He, Hong Zhang
Camera-based occupancy prediction provides geometric and semantic scene understanding for robot perception and has been adopted in autonomous vehicle and UAV perception systems. Ga…
KnowledgeBerg: Evaluating Systematic Knowledge Coverage and Compositional Reasoning in Large Language Models
Xiao Zhang, Qianru Meng, Yongjian Chen +2
Many real-world questions appear deceptively simple yet implicitly demand two capabilities: (i) systematic coverage of a bounded knowledge universe and (ii) compositional set-based…
RankSteer: Can Pointwise LLM Rankers Be Calibrated at the Representation Level?
Yumeng Wang, Catherine Chen, Suzan Verberne
Large language models (LLMs) are strong zero-shot pointwise rankers, but lag behind pairwise and listwise methods. Beyond missing comparative signals, we identify a \textit{calibra…
Benchmark Leakage Trap: Can We Trust LLM-based Recommendation?
Mingqiao Zhang, Qiyao Peng, Yinghui Wang +2
The expanding integration of Large Language Models (LLMs) into recommender systems poses critical challenges to evaluation reliability. This paper identifies and investigates a pre…
Needle in the Web: A Benchmark for Retrieving Targeted Web Pages in the Wild
Yumeng Wang, Tianyu Fan, Lingrui Xu +1
Large Language Models (LLMs) have evolved from simple chatbots into sophisticated agents capable of automating complex real-world tasks, where browsing and reasoning over live web…
How role-play shapes relevance judgment in zero-shot LLM rankers
Yumeng Wang, Jirui Qi, Catherine Chen +2
Large Language Models (LLMs) have emerged as promising zero-shot rankers, but their performance is highly sensitive to prompt formulation. In particular, role-play prompts, where t…
CoKnowledge: Supporting Assimilation of Time-synced Collective Knowledge in Online Science Videos
Yuanhao Zhang, Yumeng Wang, Xiyuan Wang +3
Danmaku, a system of scene-aligned, time-synced, floating comments, can augment video content to create 'collective knowledge'. However, its chaotic nature often hinders viewers fr…
QUIDS: Query Intent Description for Exploratory Search via Dual Space Modeling
Yumeng Wang, Xiuying Chen, Suzan Verberne
In exploratory search, users often submit vague queries to investigate unfamiliar topics, but receive limited feedback about how the search engine understood their input. This lead…
LoMA: Lossless Compressed Memory Attention
Yumeng Wang, Zhenyang Xiao
Large Language Models (LLMs) face limitations due to the high demand on GPU memory and computational resources when handling long contexts. While sparsify the Key-Value (KV) cache…