MEXA: Towards General Multimodal Reasoning with Dynamic Multi-Expert Aggregation
This work addresses the problem of scalable and flexible multimodal reasoning for applications such as medical diagnosis and financial forecasting, though it is incremental as it builds on existing expert models and aggregation techniques.
The paper tackles the challenge of building a unified framework for multimodal reasoning across diverse domains by introducing MEXA, a training-free framework that dynamically aggregates expert models, resulting in consistent performance improvements over strong baselines on benchmarks like Video Reasoning and Medical QA.
Combining pre-trained expert models offers substantial potential for scalable multimodal reasoning, but building a unified framework remains challenging due to the increasing diversity of input modalities and task complexity. For instance, medical diagnosis requires precise reasoning over structured clinical tables, while financial forecasting depends on interpreting plot-based data to make informed predictions. To tackle this challenge, we introduce MEXA, a training-free framework that performs modality- and task-aware aggregation of multiple expert models to enable effective multimodal reasoning across diverse and distinct domains. MEXA dynamically selects expert models based on the input modality and the task-specific reasoning demands (i.e., skills). Each expert model, specialized in a modality task pair, generates interpretable textual reasoning outputs. MEXA then aggregates and reasons over these outputs using a Large Reasoning Model (LRM) to produce the final answer. This modular design allows flexible and transparent multimodal reasoning across diverse domains without additional training overhead. We extensively evaluate our approach on diverse multimodal benchmarks, including Video Reasoning, Audio Reasoning, 3D Understanding, and Medical QA. MEXA consistently delivers performance improvements over strong multimodal baselines, highlighting the effectiveness and broad applicability of our expert-driven selection and aggregation in diverse multimodal reasoning tasks.