Request Correction
Use this form to request corrections to the paper metadata. Select the fields that need correction and provide the correct information.
Correction Guidelines
- Click the edit button next to a field to report a correction.
- Fill in the suggested correction value for each field you want to correct.
- Provide your name and email so we can contact you if needed.
Paper Information
Probing Multimodal Large Language Models for Global and Local Semantic Representations
Paper Fields
Click the edit button next to a field to report a correction.
Probing Multimodal Large Language Models for Global and Local Semantic Representations
The advancement of Multimodal Large Language Models (MLLMs) has greatly accelerated the development of applications in understanding integrated texts and images. Recent works leverage image-caption datasets to train MLLMs, achieving state-of-the-art performance on image-to-text tasks. However, there are few studies exploring which layers of MLLMs make the most effort to the global image information, which plays vital roles in multimodal comprehension and generation. In this study, we find that the intermediate layers of models can encode more global semantic information, whose representation vectors perform better on visual-language entailment tasks, rather than the topmost layers. We further probe models regarding local semantic representations through object recognition tasks. We find that the topmost layers may excessively focus on local information, leading to a diminished ability to encode global information. Our code and data are released via https://github.com/kobayashikanna01/probing_MLLM_rep.
Authors
Expand an author to correct their information. Use the remove button to request author removal, or add a new author.
PDF Attachment
You may attach a PDF as a corrected version of the paper. Max file size: 10MB. Only PDF files are accepted.
Your Information
Author Declaration *
Select at least one field to correct using the edit buttons above.