Exploring the Implicit Semantic Ability of Multimodal Large Language Models: A Pilot Study on Entity Set Expansion

TL;DR

This study investigates the ability of multimodal large language models to understand implicit semantic information through an entity set expansion task, introducing a novel ranking method that enhances performance.

Contribution

It is the first to apply generative MLLMs to entity set expansion and introduces the LUSAR listwise ranking method for improved semantic understanding.

Findings

LUSAR significantly improves MLLM performance on MESE.

First use of generative MLLMs for entity set expansion.

Extends listwise ranking applicability to multimodal models.

Abstract

The rapid development of multimodal large language models (MLLMs) has brought significant improvements to a wide range of tasks in real-world applications. However, LLMs still exhibit certain limitations in extracting implicit semantic information. In this paper, we apply MLLMs to the Multi-modal Entity Set Expansion (MESE) task, which aims to expand a handful of seed entities with new entities belonging to the same semantic class, and multi-modal information is provided with each entity. We explore the capabilities of MLLMs to understand implicit semantic information at the entity-level granularity through the MESE task, introducing a listwise ranking method LUSAR that maps local scores to global rankings. Our LUSAR demonstrates significant improvements in MLLM's performance on the MESE task, marking the first use of generative MLLM for ESE tasks and extending the applicability of listwise ranking.