SHINE: A Scalable In-Context Hypernetwork for Mapping Context to LoRA in a Single Pass

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Liu, Yewei, Wang, Xiyuan, Mao, Yansheng, Gelbery, Yoav, Maron, Haggai, Zhang, Muhan
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917516161318912
author Liu, Yewei
Wang, Xiyuan
Mao, Yansheng
Gelbery, Yoav
Maron, Haggai
Zhang, Muhan
author_facet Liu, Yewei
Wang, Xiyuan
Mao, Yansheng
Gelbery, Yoav
Maron, Haggai
Zhang, Muhan
contents We propose SHINE (Scalable Hyper In-context NEtwork), a scalable hypernetwork that can map diverse meaningful contexts into high-quality LoRA adapters for large language models (LLMs). By reusing the frozen LLM's own parameters in an in-context hypernetwork design and introducing architectural innovations, SHINE overcomes key limitations of prior hypernetworks and achieves strong expressive power with a relatively small number of parameters. We introduce a pretraining and instruction fine-tuning pipeline, and train our hypernetwork to generate high quality LoRA adapters from diverse meaningful contexts in a single forward pass. It updates LLM parameters without any fine-tuning, and immediately enables complex question answering tasks related to the context without directly accessing the context, effectively transforming in-context knowledge to in-parameter knowledge in one pass. Our work achieves outstanding results on various tasks, greatly saves time, computation and memory costs compared to SFT-based LLM adaptation, and shows great potential for scaling. Our code is available at https://github.com/MuLabPKU/SHINE
format Preprint
id arxiv_https___arxiv_org_abs_2602_06358
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SHINE: A Scalable In-Context Hypernetwork for Mapping Context to LoRA in a Single Pass
Liu, Yewei
Wang, Xiyuan
Mao, Yansheng
Gelbery, Yoav
Maron, Haggai
Zhang, Muhan
Computation and Language
Artificial Intelligence
We propose SHINE (Scalable Hyper In-context NEtwork), a scalable hypernetwork that can map diverse meaningful contexts into high-quality LoRA adapters for large language models (LLMs). By reusing the frozen LLM's own parameters in an in-context hypernetwork design and introducing architectural innovations, SHINE overcomes key limitations of prior hypernetworks and achieves strong expressive power with a relatively small number of parameters. We introduce a pretraining and instruction fine-tuning pipeline, and train our hypernetwork to generate high quality LoRA adapters from diverse meaningful contexts in a single forward pass. It updates LLM parameters without any fine-tuning, and immediately enables complex question answering tasks related to the context without directly accessing the context, effectively transforming in-context knowledge to in-parameter knowledge in one pass. Our work achieves outstanding results on various tasks, greatly saves time, computation and memory costs compared to SFT-based LLM adaptation, and shows great potential for scaling. Our code is available at https://github.com/MuLabPKU/SHINE
title SHINE: A Scalable In-Context Hypernetwork for Mapping Context to LoRA in a Single Pass
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2602.06358