![[2602.20624] Physics-based phenomenological characterization of cross-modal bias in multimodal models](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.20624] Physics-based phenomenological characterization of cross-modal bias in multimodal models
Summary
This paper explores the cross-modal bias in multimodal large language models (MLLMs) through a physics-based phenomenological approach, aiming to enhance algorithmic fairness by analyzing the dynamics of model interactions.
Why It Matters
Understanding cross-modal bias is crucial for improving the fairness and accuracy of AI models, especially as MLLMs become more prevalent. This research provides a novel framework that could lead to better interpretability and mitigation strategies for biases in AI systems.
Key Takeaways
- Cross-modal bias in MLLMs can lead to systematic inaccuracies.
- A physics-based approach offers new insights into model dynamics.
- Phenomenological methods can enhance understanding of algorithmic fairness.
Computer Science > Artificial Intelligence arXiv:2602.20624 (cs) [Submitted on 24 Feb 2026] Title:Physics-based phenomenological characterization of cross-modal bias in multimodal models Authors:Hyeongmo Kim, Sohyun Kang, Yerin Choi, Seungyeon Ji, Junhyuk Woo, Hyunsuk Chung, Soyeon Caren Han, Kyungreem Han View a PDF of the paper titled Physics-based phenomenological characterization of cross-modal bias in multimodal models, by Hyeongmo Kim and 7 other authors View PDF HTML (experimental) Abstract:The term 'algorithmic fairness' is used to evaluate whether AI models operate fairly in both comparative (where fairness is understood as formal equality, such as "treat like cases as like") and non-comparative (where unfairness arises from the model's inaccuracy, arbitrariness, or inscrutability) contexts. Recent advances in multimodal large language models (MLLMs) are breaking new ground in multimodal understanding, reasoning, and generation; however, we argue that inconspicuous distortions arising from complex multimodal interaction dynamics can lead to systematic bias. The purpose of this position paper is twofold: first, it is intended to acquaint AI researchers with phenomenological explainable approaches that rely on the physical entities that the machine experiences during training/inference, as opposed to the traditional cognitivist symbolic account or metaphysical approaches; second, it is to state that this phenomenological doctrine will be practically useful for tackl...
