Can I read Self-Improvement in Multimodal Large Language Models: A Survey on EtoBox?

Self-Improvement in Multimodal Large Language Models: A Survey by Shijian Deng & Kai Wang & Tianyu Yang & Harsh Singh & Yapeng Tian is a book available to read on EtoBox.

What is Self-Improvement in Multimodal Large Language Models: A Survey about?

AbstractarXiv:2510.02665v1 [cs.CL] 3 Oct 2025Recent advancements in self-improvement forLarge Language Models (LLMs) have efficiently enhanced model capabilities withoutsignificantly increasing costs, particularly interms of human effort. While this area isstill relatively young, its extension to the multimodal domain holds immense potential forleveraging diverse data sources and developing more general self-improving models. Thissurvey is the first to provide a comprehensiveFiguroverview of self-improvement in MultimodalmodaLLMs (MLLMs). We provide a structuredselectoverview of the current literature and discussit intomethods from three perspectives: 1) data colcolleclection, 2) data organization, and 3) model opthroutimization, to facilitate the further developmentrecurof self-improvement in MLLMs. We also include commonly used evaluations and downstream applications. Finally, we conclude byoutlining open challenges and future researchsingldirections.self-iet al.1 Introductionin texeralSelf-improvement aims to enable models to colet allect and organize data required to build a betternatiogeneration of themselves, which offers a path tosive sovercome the costly scaling iss

Author

Shijian Deng & Kai Wang & Tianyu Yang & Harsh Singh & Yapeng Tian

Published

2025

Language

EN