Panza: Design and Analysis of a Fully-Local Personalized Text Writing Assistant
Armand Nicolicioiu, Eugenia Iofinova, Andrej Jovanovic, Eldar Kurtic,, Mahdi Nikdan, Andrei Panferov, Ilia Markov, Nir Shavit, Dan Alistarh

TL;DR
Panza is a personalized email writing assistant that fine-tunes large language models using limited personal data on resource-constrained devices, demonstrating effective style imitation and raising privacy concerns.
Contribution
The paper introduces a novel design combining fine-tuning and RAG for personalized email generation, with detailed evaluation metrics and analysis of system components.
Findings
Effective style imitation with under 100 emails
Fine-tuning on limited resources is feasible
Small data can enable impersonation attacks
Abstract
The availability of powerful open-source large language models (LLMs) opens exciting use-cases, such as using personal data to fine-tune these models to imitate a user's unique writing style. Two key requirements for such assistants are personalization - in the sense that the assistant should recognizably reflect the user's own writing style - and privacy - users may justifiably be wary of uploading extremely personal data, such as their email archive, to a third-party service. In this paper, we present a new design and evaluation for such an automated assistant, for the specific use case of email generation, which we call Panza. Panza's personalization features are based on a combination of fine-tuning using a variant of the Reverse Instructions technique together with Retrieval-Augmented Generation (RAG). We demonstrate that this combination allows us to fine-tune an LLM to reflect a…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsTopic Modeling · Natural Language Processing Techniques · Recommender Systems and Techniques
Methodstravel james · Attention Is All You Need · Layer Normalization · Refunds@Expedia|||How do I get a full refund from Expedia? · Dense Connections · Adam · Softmax · Linear Warmup With Linear Decay · Residual Connection · Dropout
