From GPT-4 to Gemini and Beyond: Assessing the Landscape of MLLMs on Generalizability, Trustworthiness and Causality through Four Modalities
Chaochao Lu, Chen Qian, Guodong Zheng, Hongxing Fan, Hongzhi Gao, Jie Zhang, Jing Shao, Jingyi Deng, Jinlan Fu, Kexin Huang, Kunchang Li, Lijun Li, Limin Wang, Lu Sheng, Meiqi Chen, Ming Zhang, Qibing Ren, Sirui Chen, Tao Gui, Wanli Ouyang, Yali Wang, Yan Teng, Yaru Wang, Yi Wang, Yinan He, Yingchun Wang, Yixu Wang, Yongting Zhang, Yu Qiao, Yujiong Shen, Yurong Mou, Yuxi Chen, Zaibin Zhang, Zhelun Shi, Zhenfei Yin, Zhipin Wang
TL;DR
The paper investigates generalizability, trustworthiness, and causal reasoning of multi-modal LLMs across text, code, image, and video. It benchmarks GPT-4, Gemini, and six open-source MLLMs using a uniform, qualitative evaluation of 230 hand-designed cases, condensed into 12 modality-property scores. Across four modalities, it reports 14 empirical findings highlighting strengths and limitations of proprietary versus open-source models, with notable gaps in math, domain knowledge, safety, and causal understanding. The work emphasizes transparency and standardized evaluation, proposing ongoing leaderboard updates to guide reliable downstream multi-modal applications.
Abstract
Multi-modal Large Language Models (MLLMs) have shown impressive abilities in generating reasonable responses with respect to multi-modal contents. However, there is still a wide gap between the performance of recent MLLM-based applications and the expectation of the broad public, even though the most powerful OpenAI's GPT-4 and Google's Gemini have been deployed. This paper strives to enhance understanding of the gap through the lens of a qualitative study on the generalizability, trustworthiness, and causal reasoning capabilities of recent proprietary and open-source MLLMs across four modalities: ie, text, code, image, and video, ultimately aiming to improve the transparency of MLLMs. We believe these properties are several representative factors that define the reliability of MLLMs, in supporting various downstream applications. To be specific, we evaluate the closed-source GPT-4 and Gemini and 6 open-source LLMs and MLLMs. Overall we evaluate 230 manually designed cases, where the qualitative results are then summarized into 12 scores (ie, 4 modalities times 3 properties). In total, we uncover 14 empirical findings that are useful to understand the capabilities and limitations of both proprietary and open-source MLLMs, towards more reliable downstream multi-modal applications.