Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
Improving Task-Specific Multimodal Sentiment Analysis with General MLLMs via Prompting
Authors: Haoyu Zhang, Yinan Zhang, Chaolong Ying, Xiaoying Tang, Tianshu Yu
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on the SIMS, MOSI, and MOSEI datasets demonstrate that our framework enables task-specific models to achieve state-of-the-art performance across most metrics. |
| Researcher Affiliation | Academia | Haoyu Zhang , Yinan Zhang , Chaolong Ying , Xiaoying Tang , Tianshu Yu School of Data Science, The Chinese University of Hong Kong, Shenzhen School of Science and Engineering, The Chiese University of Hong Kong, Shenzhen EMAIL EMAIL |
| Pseudocode | No | The paper describes the methodology using prose, mathematical equations, and diagrams, but does not include any clearly labeled pseudocode or algorithm blocks. |
| Open Source Code | Yes | Correspondence author 1Code: https://github.com/LOGO-CUHKSZ/MMSLF |
| Open Datasets | Yes | Extensive experiments on the SIMS, MOSI, and MOSEI datasets demonstrate that our framework enables task-specific models to achieve state-of-the-art performance across most metrics. (In NeurIPS Paper Checklist, item 5: The datasets used in the paper are open datasets and can be accessed by anyone upon request.) |
| Dataset Splits | Yes | SIMS. SIMS [13] is a Chinese MSA dataset... It consists of 1,368 training samples, 456 validation samples, and 457 test samples. MOSI. MOSI [14] is an English MSA dataset... The dataset includes 1,284 training samples, 229 validation samples, and 686 test samples. MOSEI. MOSEI [15] is an English MSA dataset... It contains 22,856 video clips, including 16,326 training samples, 1,871 validation samples, and 4,659 test samples. |
| Hardware Specification | Yes | The experiments were conducted on a PC equipped with an AMD EPYC 7513 processor (2.6GHz) and an NVIDIA Tesla A40 GPU. |
| Software Dependencies | Yes | We implemented our proposed method using Py Torch 2.1.1 with CUDA 12.1. |
| Experiment Setup | Yes | The key parameters are listed in Table 20. (Table 20 contains detailed parameters such as Batch Size, Optimizer, Epochs, Seeds, Initial Learning Rate, and regularization weights α, β for SIMS, MOSI, and MOSEI datasets.) |