Saved in:
Bibliographic Details
Main Authors: Li, Xiaoyu, Liu, Yuhang, Kang, Xuanshuo, Luo, Zheng, Lou, Fangqi, Wu, Xiaohua, Xiong, Zihan
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.12760
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911547859664896
author Li, Xiaoyu
Liu, Yuhang
Kang, Xuanshuo
Luo, Zheng
Lou, Fangqi
Wu, Xiaohua
Xiong, Zihan
author_facet Li, Xiaoyu
Liu, Yuhang
Kang, Xuanshuo
Luo, Zheng
Lou, Fangqi
Wu, Xiaohua
Xiong, Zihan
contents In-Context Learning (ICL) is a significant paradigm for Large Multimodal Models (LMMs), using a few in-context demonstrations (ICDs) for new task adaptation. However, its performance is sensitive to demonstration configurations and computationally expensive. Mathematically, the influence of these demonstrations can be decomposed into a dynamic mixture of the standard attention output and the context values. Current approximation methods simplify this process by learning a "shift vector". Inspired by the exact decomposition, we introduce High-Fidelity In-Context Learning (HIFICL) to more faithfully model the ICL mechanism. HIFICL consists of three key components: 1) a set of "virtual key-value pairs" to act as a learnable context, 2) a low-rank factorization for stable and regularized training, and 3) a simple end-to-end training objective. From another perspective, this mechanism constitutes a form of context-aware Parameter-Efficient Fine-Tuning (PEFT). Extensive experiments show that HiFICL consistently outperforms existing approximation methods on several multimodal benchmarks. The code is available at https://github.com/bbbandari/HiFICL.
format Preprint
id arxiv_https___arxiv_org_abs_2603_12760
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle HIFICL: High-Fidelity In-Context Learning for Multimodal Tasks
Li, Xiaoyu
Liu, Yuhang
Kang, Xuanshuo
Luo, Zheng
Lou, Fangqi
Wu, Xiaohua
Xiong, Zihan
Computer Vision and Pattern Recognition
In-Context Learning (ICL) is a significant paradigm for Large Multimodal Models (LMMs), using a few in-context demonstrations (ICDs) for new task adaptation. However, its performance is sensitive to demonstration configurations and computationally expensive. Mathematically, the influence of these demonstrations can be decomposed into a dynamic mixture of the standard attention output and the context values. Current approximation methods simplify this process by learning a "shift vector". Inspired by the exact decomposition, we introduce High-Fidelity In-Context Learning (HIFICL) to more faithfully model the ICL mechanism. HIFICL consists of three key components: 1) a set of "virtual key-value pairs" to act as a learnable context, 2) a low-rank factorization for stable and regularized training, and 3) a simple end-to-end training objective. From another perspective, this mechanism constitutes a form of context-aware Parameter-Efficient Fine-Tuning (PEFT). Extensive experiments show that HiFICL consistently outperforms existing approximation methods on several multimodal benchmarks. The code is available at https://github.com/bbbandari/HiFICL.
title HIFICL: High-Fidelity In-Context Learning for Multimodal Tasks
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.12760