7 papers
The Matching Number Characterizes the Query Complexity of Edge Counting and Degree Moment Estimation
Chenglin Fan, Wonseok Shin
Consider an unknown graph with vertex degrees , accessed only by sampling uniformly random edges and querying the degrees of their endpoints. We study the query co…
Visual Information-Guided Parallel Decoding for Diffusion Multimodal Large Language Models
Insu Lee, Wooje Park, Wonseok Shin +2
Diffusion multimodal large language models (dMLLMs) have recently emerged as a new decoding paradigm for multimodal generation. Starting from a fully masked sequence, dMLLMs progre…
Aligning Sentence Embeddings to Human Concepts via Sparse Autoencoders
Wonseok Shin, Songkuk Kim
Dense sentence embeddings are fundamental to modern Retrieval-Augmented Generation (RAG) systems but suffer from a lack of interpretability due to feature superposition. This opaci…
Sparsity as a Key: Unlocking New Insights from Latent Structures for Out-of-Distribution Detection
Ahyoung Oh, Wonseok Shin, Songkuk Kim
Sparse Autoencoders (SAEs) have demonstrated significant success in interpreting Large Language Models (LLMs) by decomposing dense representations into sparse, semantic components.…
Efficient Hypergraph Pattern Matching via Match-and-Filter and Intersection Constraint
Siwoo Song, Wonseok Shin, Kunsoo Park +3
A hypergraph is a generalization of a graph, in which a hyperedge can connect multiple vertices, modeling complex relationships involving multiple vertices simultaneously. Hypergra…
Cardinality Estimation of Subgraph Matching: A Filtering-Sampling Approach
Wonseok Shin, Siwoo Song, Kunsoo Park +1
Subgraph counting is a fundamental problem in understanding and analyzing graph structured data, yet computationally challenging. This calls for an accurate and efficient algorithm…
Triangular Contrastive Learning on Molecular Graphs
MinGyu Choi, Wonseok Shin, Yijingxiu Lu +1
Recent contrastive learning methods have shown to be effective in various tasks, learning generalizable representations invariant to data augmentation thereby leading to state of t…