LLM-Augmented Semantic Steering of Text Embedding Projection Spaces
This work addresses the problem of inflexible text embedding projections for visual document analysis, offering a more interpretable and flexible semantic interaction method.
The paper introduces LLM-augmented semantic steering, which allows analysts to reorganize text embedding projections by grouping example documents and using an LLM to extend this intent to related documents via text augmentation or embedding blending. Simulation-based evaluation shows improved global and local alignment with target semantic structures using minimal interaction.
Low-dimensional projections of text embeddings support 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 introduce LLM-augmented semantic steering, which enables analysts to express semantic intent by grouping a small set of example documents within the projection. A large language model externalizes this intent as natural-language representations and selectively extends it to related documents; the resulting semantic information is then incorporated into document representations via text augmentation or embedding-level blending, without retraining the underlying models. A case study illustrates how the same corpus can be reorganized from different semantic perspectives, while simulation-based evaluation shows that semantic steering improves global and local alignment with target semantic structures using only minimal interaction. Embedding-level blending further enables continuous and controllable steering of projection layouts. These results position projection spaces as intent-dependent semantic workspaces that can be reshaped through explicit, interpretable, language-mediated interaction.