Intrinsic explainability of multimodal learning for crop yield simulation
Studies how multimodal crop-yield models can be made more interpretable, examining modality contributions and model behavior rather than treating multimodal fusion as a black box.
h.-najjar