AI RESEARCH
LLM-Augmented Semantic Steering of Text Embedding Projection Spaces
arXiv CS.CL
•
ArXi:2605.01957v1 Announce Type: cross Low-dimensional projections of text embeddings visual analysis of document collections, but their spatial organization may not reflect the relationships an analyst intends to examine. Existing semantic interaction approaches encode semantic intent indirectly through geometric constraints or model updates, limiting interpretability and flexibility. We