5 papers
PromptEmbedder: Efficient and Transferable Text Embedding via Dual-LLM Soft Prompting
Yu-Che Tsai, Kuan-Yu Chen, Yuan-Hao Chen +4
Large Language Models (LLMs) have demonstrated remarkable efficacy in text embedding, yet current adaptation methods like LoRA face significant bottlenecks in computational efficie…
Decision-aware User Simulation Agent for Evaluating Conversational Recommender Systems
Yuan-Chi Li, Li-Chi Chen, Sung-Yi Wu +2
Conversational recommender systems (CRS) increasingly rely on user simulators for automated evaluation of sales agents. A key requirement for such simulators is the ability to mode…
Concept-Aware Privacy Mechanisms for Defending Embedding Inversion Attacks
Yu-Che Tsai, Hsiang Hsiao, Kuan-Yu Chen +1
Text embeddings enable numerous NLP applications but face severe privacy risks from embedding inversion attacks, which can expose sensitive attributes or reconstruct raw text. Exis…
Let LLMs Speak Embedding Languages: Generative Text Embeddings via Iterative Contrastive Refinement
Yu-Che Tsai, Kuan-Yu Chen, Yuan-Chi Li +3
Existing large language model (LLM)-based embeddings typically adopt an encoder-only paradigm, treating LLMs as static feature extractors and overlooking their core generative stre…
Transferable Embedding Inversion Attack: Uncovering Privacy Risks in Text Embeddings without Model Queries
Yu-Hsiang Huang, Yuche Tsai, Hsiang Hsiao +2
This study investigates the privacy risks associated with text embeddings, focusing on the scenario where attackers cannot access the original embedding model. Contrary to previous…