Do Implicit Personalization and Explicit Styles Conflict? PsPLUG: A Lightweight Plug-in for Balancing Personalization and Style in Customized LLMs
For developers of personalized LLMs, this work addresses a practical failure mode where style instructions degrade personalization, offering a simple fix.
The paper identifies that explicit style instructions can conflict with implicit user preferences in personalized LLMs, causing personalization collapse. PsPLUG, a lightweight plug-in, resolves this by learning user-specific residuals after accounting for style, enabling better preservation of user preferences while maintaining style control.
Personalized large language models are often expected to follow explicit style instructions, yet we find that such instructions can undermine the user-specific characteristics that personalization methods aim to preserve. We call this failure mode personalization collapse: explicit style control can conflict with implicit user preferences. To address this challenge, we propose PsPLUG, a lightweight plug-in that learns a user-specific residual after accounting for the requested style. PsPLUG also allows us to tune personalization strength at inference time. Our experiments show that explicit style instructions can diminish personalization in existing methods, whereas PsPLUG better preserves user preferences while providing precise control over the balance between personalization and style adherence.