# Enabling On-Device Large Language Model Personalization with Self-Supervised Data Selection and Synthesis

arXiv: https://arxiv.org/abs/2311.12275 · HTML: https://mlsanity.com/paper/2311.12275/

## Key facts

- Cited by 20 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 29 references that are in the map.
- First posted on arXiv 21 Nov 2023.

## Builds on (most cited first, 17 of 17)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Finetuned Language Models Are Zero-Shot Learners](https://mlsanity.com/paper/2109.01652/) (2021)
- [RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control](https://mlsanity.com/paper/2307.15818/) (2023)
- [Gradient based sample selection for online continual learning](https://mlsanity.com/paper/1903.08671/) (2019)
- [Coresets via Bilevel Optimization for Continual Learning and Streaming](https://mlsanity.com/paper/2006.03875/) (2020)
- [Word Alignment by Fine-tuning Embeddings on Parallel Corpora](https://mlsanity.com/paper/2101.08231/) (2021)
- [MedDialog: Two Large-scale Medical Dialogue Datasets](https://mlsanity.com/paper/2004.03329/) (2020)
- [Memory Efficient Experience Replay for Streaming Learning](https://mlsanity.com/paper/1809.05922/) (2018)
- [ProsocialDialog: A Prosocial Backbone for Conversational Agents](https://mlsanity.com/paper/2205.12688/) (2022)
- [Federated Learning of Gboard Language Models with Differential Privacy](https://mlsanity.com/paper/2305.18465/) (2023)
- [Question Answering as an Automatic Evaluation Metric for News Article Summarization](https://mlsanity.com/paper/1906.00318/) (2019)
- [Self-supervised learning through the eyes of a child](https://mlsanity.com/paper/2007.16189/) (2020)
- [Enabling On-Device Self-Supervised Contrastive Learning With Selective Data Contrast](https://mlsanity.com/paper/2106.03796/) (2021)
- [When Automated Assessment Meets Automated Content Generation: Examining Text Quality in the Era of GPTs](https://mlsanity.com/paper/2309.14488/) (2023)
- [Self-supervised On-device Federated Learning from Unlabeled Streams](https://mlsanity.com/paper/2212.01006/) (2022)
- [IBERT: Idiom Cloze-style reading comprehension with Attention](https://mlsanity.com/paper/2112.02994/) (2021)

## Built on by (most cited first, 20 of 20)

- [From Persona to Personalization: A Survey on Role-Playing Language Agents](https://mlsanity.com/paper/2404.18231/) (2024)
- [Two Tales of Persona in LLMs: A Survey of Role-Playing and Personalization](https://mlsanity.com/paper/2406.01171/) (2024)
- [A Survey of Personalized Large Language Models: Progress and Future Directions](https://mlsanity.com/paper/2502.11528/) (2025)
- [The Rise of Small Language Models in Healthcare: A Comprehensive Survey](https://mlsanity.com/paper/2504.17119/) (2025)
- [Empirical Guidelines for Deploying LLMs onto Resource-constrained Edge Devices](https://mlsanity.com/paper/2406.03777/) (2024)
- [Robust Implementation of Retrieval-Augmented Generation on Edge-based Computing-in-Memory Architectures](https://mlsanity.com/paper/2405.04700/) (2024)
- [Enabling On-Device LLMs Personalization with Smartphone Sensing](https://mlsanity.com/paper/2407.04418/) (2024)
- [A Survey on Collaborating Small and Large Language Models for Performance, Cost-effectiveness, Cloud-edge Privacy, and Trustworthiness](https://mlsanity.com/paper/2510.13890/) (2025)
- [Thinking Before Speaking: A Role-playing Model with Mindset](https://mlsanity.com/paper/2409.13752/) (2024)
- [Adaptive Self-Supervised Learning Strategies for Dynamic On-Device LLM Personalization](https://mlsanity.com/paper/2409.16973/) (2024)
- [Are We There Yet? A Measurement Study of Efficiency for LLM Applications on Mobile Devices](https://mlsanity.com/paper/2504.00002/) (2025)
- [Personalizing Large Language Models using Retrieval Augmented Generation and Knowledge Graph](https://mlsanity.com/paper/2505.09945/) (2025)
- [Memory- and Latency-Constrained Inference of Large Language Models via Adaptive Split Computing](https://mlsanity.com/paper/2511.04002/) (2025)
- [Tiny-Align: Bridging Automatic Speech Recognition and Large Language Model on the Edge](https://mlsanity.com/paper/2411.13766/) (2024)
- [Prada: Black-Box LLM Adaptation with Private Data on Resource-Constrained Devices](https://mlsanity.com/paper/2503.14932/) (2025)
- [State Your Intention to Steer Your Attention: An AI Assistant for Intentional Digital Living](https://mlsanity.com/paper/2510.14513/) (2025)
- [NVCiM-PT: An NVCiM-assisted Prompt Tuning Framework for Edge LLMs](https://mlsanity.com/paper/2411.08244/) (2024)
- [TSUBASA: Improving Long-Horizon Personalization via Evolving Memory and Self-Learning with Context Distillation](https://mlsanity.com/paper/2604.07894/) (2026)
- [An Adaptive System for Wearable Devices to Detect Stress Using Physiological Signals](https://mlsanity.com/paper/2407.15252/) (2024)
- [MemSpec: Memory-Aware Runtime for Adaptive Draft Scheduling in Speculative Decoding on Edge Devices](https://mlsanity.com/paper/2608.10362/) (2026)

## Abstract (arXiv)

After a large language model (LLM) is deployed on edge devices, it is desirable for these devices to learn from user-generated conversation data to generate user-specific and personalized responses in real-time. However, user-generated data usually contains sensitive and private information, and uploading such data to the cloud for annotation is not preferred if not prohibited. While it is possible to obtain annotation locally by directly asking users to provide preferred responses, such annotations have to be sparse to not affect user experience. In addition, the storage of edge devices is usually too limited to enable large-scale fine-tuning with full user-generated data. It remains an open question how to enable on-device LLM personalization, considering sparse annotation and limited on-device storage. In this paper, we propose a novel framework to select and store the most representative data online in a self-supervised way. Such data has a small memory footprint and allows infrequent requests of user annotations for further fine-tuning. To enhance fine-tuning quality, multiple semantically similar pairs of question texts and expected responses are generated using the LLM. Our experiments show that the proposed framework achieves the best user-specific content-generating capability (accuracy) and fine-tuning speed (performance) compared with vanilla baselines. To the best of our knowledge, this is the very first on-device LLM personalization framework.