Proposes hypergraph-based dynamic topic modeling using structured low-rank factorizations, temporal regularization, and a nonlinear multinomial likelihood, with local convergence guarantees and error bounds.
Hongjie Cai, Rui Xia, and Jianfei Yu
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
HyPE encodes personas as category-induced hypergraphs with persistent edge embeddings and reports consistent gains over sentence pooling on PersonaChat for GPT-2, LLaMA-3.2-3B and Qwen2.5-3B under greedy decoding.
citing papers explorer
-
Dynamic Topic Modeling with a Higher-Order Hypergraphical Representation
Proposes hypergraph-based dynamic topic modeling using structured low-rank factorizations, temporal regularization, and a nonlinear multinomial likelihood, with local convergence guarantees and error bounds.
-
HyPE: Category-Aware Hypergraph Encoding with Persistent Edge Embeddings for Persona-Grounded Dialogue
HyPE encodes personas as category-induced hypergraphs with persistent edge embeddings and reports consistent gains over sentence pooling on PersonaChat for GPT-2, LLaMA-3.2-3B and Qwen2.5-3B under greedy decoding.