Work

CoGR-MoE: Concept-Guided Expert Routing with Consistent Selection and Flexible Reasoning for Visual Question Answering

Year: 2026

Type: conference-paper

Abstract: Visual Question Answering (VQA) requires models to identify the correct answer options based on both visual and textual evidence.Recent Mixture-of-Experts (MoE) methods improve option reasoning by grouping similar concepts or routing based on examples.However, unstable routing can lead to inconsistent expert selection in the same question type, while overly stable routing may reduce flexibility.To address this, we propose Concept-Guided Routing framework (CoGR-MoE), which incorporates semantics of the answer options to guide expert selection in the training phase.Next, option features are used to reweight the selected experts, producing discriminative representations for each candidate option.These optionlevel representations are further used for option comparison and optimized via contrastive learning.The experimental results indicate that CoGR-MoE delivers strong performance across multiple VQA tasks, demonstrating the effectiveness of our approach.Our code is available at CoGR-MoE. Option-Aware ReweightingRouter T … (I, Q) a. Cornish Cat large ears broad face b.Devon Rex low ears narrow face c.Sphnyx hairless skin small ears d.Selkirk Rex wide eyes sparse coat What breed is this cat?

Source: Findings of the Association for Computational Linguistics: ACL (Findings of ACL)

DOI: 10.18653/v1/2026.findings-acl.315

Authors: Xiyin Zeng, Yi Lu, Hao Wang

Institution: The Hong Kong University of Science and Technology (Guangzhou)

Language: English

Topic: Multimodal Machine Learning Applications

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

Funder: Guangzhou Municipal Science and Technology Project

OpenAlex ID: W7166802212