Embedding-to-Prefix: Parameter-Efficient Personalization for Pre-Trained Large Language Models
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916754226151424 |
|---|---|
| author | Huber, Bernd Fazelnia, Ghazal Damianou, Andreas Peleato, Sebastian Lefarov, Max Ravichandran, Praveen De Nadai, Marco Lalmas-Roellke, Mounia Bennett, Paul N. |
| author_facet | Huber, Bernd Fazelnia, Ghazal Damianou, Andreas Peleato, Sebastian Lefarov, Max Ravichandran, Praveen De Nadai, Marco Lalmas-Roellke, Mounia Bennett, Paul N. |
| contents | Large language models (LLMs) excel at generating contextually relevant content. However, tailoring these outputs to individual users for effective personalization is a significant challenge. While rich user-specific information often exists as pre-existing user representations, such as embeddings learned from preferences or behaviors, current methods to leverage these for LLM personalization typically require costly fine-tuning or token-heavy prompting. We propose Embedding-to-Prefix (E2P), a parameter-efficient method that injects pre-computed context embeddings into an LLM's hidden representation space through a learned projection to a single soft token prefix. This enables effective personalization while keeping the backbone model frozen and avoiding expensive adaptation techniques. We evaluate E2P across two public datasets and in a production setting: dialogue personalization on Persona-Chat, contextual headline generation on PENS, and large-scale personalization for music and podcast consumption. Results show that E2P preserves contextual signals and achieves strong performance with minimal computational overhead, offering a scalable, efficient solution for contextualizing generative AI systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_17051 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Embedding-to-Prefix: Parameter-Efficient Personalization for Pre-Trained Large Language Models Huber, Bernd Fazelnia, Ghazal Damianou, Andreas Peleato, Sebastian Lefarov, Max Ravichandran, Praveen De Nadai, Marco Lalmas-Roellke, Mounia Bennett, Paul N. Computation and Language Artificial Intelligence Large language models (LLMs) excel at generating contextually relevant content. However, tailoring these outputs to individual users for effective personalization is a significant challenge. While rich user-specific information often exists as pre-existing user representations, such as embeddings learned from preferences or behaviors, current methods to leverage these for LLM personalization typically require costly fine-tuning or token-heavy prompting. We propose Embedding-to-Prefix (E2P), a parameter-efficient method that injects pre-computed context embeddings into an LLM's hidden representation space through a learned projection to a single soft token prefix. This enables effective personalization while keeping the backbone model frozen and avoiding expensive adaptation techniques. We evaluate E2P across two public datasets and in a production setting: dialogue personalization on Persona-Chat, contextual headline generation on PENS, and large-scale personalization for music and podcast consumption. Results show that E2P preserves contextual signals and achieves strong performance with minimal computational overhead, offering a scalable, efficient solution for contextualizing generative AI systems. |
| title | Embedding-to-Prefix: Parameter-Efficient Personalization for Pre-Trained Large Language Models |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2505.17051 |