SAME: Stabilized Mixture-of-Experts for Multimodal Continual Instruction Tuning

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Xie, Zhen-Hao, Tang, Jun-Tao, Shi, Yu-Cheng, Ye, Han-Jia, Zhan, De-Chuan, Zhou, Da-Wei
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910265645203456
author Xie, Zhen-Hao
Tang, Jun-Tao
Shi, Yu-Cheng
Ye, Han-Jia
Zhan, De-Chuan
Zhou, Da-Wei
author_facet Xie, Zhen-Hao
Tang, Jun-Tao
Shi, Yu-Cheng
Ye, Han-Jia
Zhan, De-Chuan
Zhou, Da-Wei
contents Multimodal Large Language Models (MLLMs) achieve strong performance through instruction tuning, but real-world deployment requires them to continually expand their capabilities, making Multimodal Continual Instruction Tuning (MCIT) essential. Recent methods leverage sparse expert routing to promote task specialization, but we find that the expert routing process suffers from drift as the data distribution evolves. For example, a grounding query that previously activated localization experts may instead be routed to irrelevant experts after learning OCR tasks. Meanwhile, the grounding-related experts can be overwritten by new tasks and lose their original functionality. Such failure reflects two problems: router drift, where expert selection becomes inconsistent over time, and expert drift, where shared experts are overwritten across tasks. Therefore, we propose StAbilized Mixture-of-Experts (SAME) for MCIT. To address router drift, SAME stabilizes expert selection by decomposing routing dynamics into orthogonal subspaces and updating only task-relevant directions. To mitigate expert drift, we regulate expert updates via curvature-aware scaling using historical input covariance in a rehearsal-free manner. SAME also introduces adaptive expert activation to freeze selected experts during training, reducing redundant computation and cross-task interference. We also introduce a new benchmark to evaluate MCIT with long task sequence, and extensive experiments demonstrate SAME's SOTA performance. Code is available at https://github.com/LAMDA-CL/Prism.
format Preprint
id arxiv_https___arxiv_org_abs_2602_01990
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SAME: Stabilized Mixture-of-Experts for Multimodal Continual Instruction Tuning
Xie, Zhen-Hao
Tang, Jun-Tao
Shi, Yu-Cheng
Ye, Han-Jia
Zhan, De-Chuan
Zhou, Da-Wei
Machine Learning
Artificial Intelligence
Multimodal Large Language Models (MLLMs) achieve strong performance through instruction tuning, but real-world deployment requires them to continually expand their capabilities, making Multimodal Continual Instruction Tuning (MCIT) essential. Recent methods leverage sparse expert routing to promote task specialization, but we find that the expert routing process suffers from drift as the data distribution evolves. For example, a grounding query that previously activated localization experts may instead be routed to irrelevant experts after learning OCR tasks. Meanwhile, the grounding-related experts can be overwritten by new tasks and lose their original functionality. Such failure reflects two problems: router drift, where expert selection becomes inconsistent over time, and expert drift, where shared experts are overwritten across tasks. Therefore, we propose StAbilized Mixture-of-Experts (SAME) for MCIT. To address router drift, SAME stabilizes expert selection by decomposing routing dynamics into orthogonal subspaces and updating only task-relevant directions. To mitigate expert drift, we regulate expert updates via curvature-aware scaling using historical input covariance in a rehearsal-free manner. SAME also introduces adaptive expert activation to freeze selected experts during training, reducing redundant computation and cross-task interference. We also introduce a new benchmark to evaluate MCIT with long task sequence, and extensive experiments demonstrate SAME's SOTA performance. Code is available at https://github.com/LAMDA-CL/Prism.
title SAME: Stabilized Mixture-of-Experts for Multimodal Continual Instruction Tuning
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2602.01990