GrootVL: Tree Topology is All You Need in State Space Model
This work addresses a key bottleneck in state space models for AI researchers and practitioners, offering a novel method to enhance long-range dependency modeling in multimodal applications.
The authors tackled the problem of modeling long-range dependencies in state space models by proposing GrootVL, which uses dynamically generated tree topologies to break sequence constraints, resulting in significant performance improvements over existing models in visual tasks and consistent gains in textual tasks with minor training cost.
The state space models, employing recursively propagated features, demonstrate strong representation capabilities comparable to Transformer models and superior efficiency. However, constrained by the inherent geometric constraints of sequences, it still falls short in modeling long-range dependencies. To address this issue, we propose the GrootVL network, which first dynamically generates a tree topology based on spatial relationships and input features. Then, feature propagation is performed based on this graph, thereby breaking the original sequence constraints to achieve stronger representation capabilities. Additionally, we introduce a linear complexity dynamic programming algorithm to enhance long-range interactions without increasing computational cost. GrootVL is a versatile multimodal framework that can be applied to both visual and textual tasks. Extensive experiments demonstrate that our method significantly outperforms existing structured state space models on image classification, object detection and segmentation. Besides, by fine-tuning large language models, our approach achieves consistent improvements in multiple textual tasks at minor training cost.