Source Themes

Lost in Decoding? Reproducing and Stress-Testing the Look-Ahead Prior in Generative Retrieval
Generative retrieval (GR) ranks documents by autoregressively generating identifiers. Trie-constrained beam search, which is used by …
Kidist Amde Mekonnen, Yongkang Li, Yubao Tang, Simon Lupart, Maarten De Rijke
On the Reproducibility of “Hypencoder: Hypernetworks for Information Retrieval”
The Hypencoder, proposed by Killingback et al. (SIGIR 2025), is a retrieval framework that replaces the fixed inner-product scoring …
Arne Eichholtz, Yongkang Li, Jutte Vijverberg, Tobias Groot, Mohammad Aliannejadi
Spectral Tempering for Embedding Compression in Dense Passage Retrieval
Dimensionality reduction is critical for deploying dense retrieval systems at scale, yet mainstream post-hoc methods face a fundamental …
Yongkang Li, Panagiotis Eustratiadis, Evangelos Kanoulas
PDF Code
Less is More: Adaptive Prompt Compression and Exemplar Selection for Efficient Few-Shot Sentiment Analysis
Few-shot sentiment analysis remains challenging due to limited labeled data and inefficient prompt design for large language models …
Peter Atandoh, Yongkang Li, Weikang Guo, Jinyu Guo, Jie Zou
PDF DOI
Multifaceted Scenario-Aware Hypergraph Learning for Next POI Recommendation
Among the diverse services provided by Location-Based Social Networks (LBSNs), Next Point-of-Interest (POI) recommendation plays a …
Yuxi Lin, Yongkang Li, Jie Xing, Zipei Fan
PDF Code DOI
Rethinking the Privacy of Text Embeddings: A Reproducibility Study of “Text Embeddings Reveal (Almost) As Much As Text”
Text embeddings are fundamental to many natural language processing~(NLP) tasks, extensively applied in domains such as recommendation …
Dominykas Seputis, Yongkang Li, Karsten Langerak, Serghei Mihailov
PDF Code DOI
Unsupervised Corpus Poisoning Attacks in Continuous Space for Dense Retrieval
This paper concerns corpus poisoning attacks in dense information retrieval, where an adversary attempts to compromise the ranking …
Yongkang Li, Panagiotis Eustratiadis,Simon Lupart, Evangelos Kanoulas
PDF Code DOI
Reproducing HotFlip for Corpus Poisoning Attacks in Dense Retrieval
HotFlip is a topical gradient-based word substitution method for attacking language models. Recently, this method has been further …
Yongkang Li, Panagiotis Eustratiadis, Evangelos Kanoulas
PDF Code DOI
Heterogeneous Hyperbolic Hypergraph Neural Network for Friend Recommendation in Location-based Social Networks
Friend recommendation is an important real-world application in location-based social networks (LBSN), helping users discover potential …
Yongkang Li, Zipei Fan, Xuan Song
PDF Code DOI
Hyper-relational knowledge graph neural network for next POI recommendation
With the advancement of mobile technology, Point of Interest (POI) recommendation systems in Location-based Social Networks (LBSN) have …
Jixiao Zhang, Yongkang Li, Ruotong Zou, Jingyuan Zhang,Renhe Jiang, Zipei Fan, Xuan Song
PDF Code DOI