Multimodal Large Language Models (MLLMs) have demonstrated outstanding performance across a variety of domains. However, training MLLMs is often inefficient due to the significantly longer input sequences introduced by multimodal data and the low utilization of inter-layer computations. To address this challenge, we shift the focus to the training process itself and propose a novel training-efficient framework based on sparse representations, termed the Sparse Training Scheme (STS). This scheme consists of two key components: the Visual Token Compressor, which reduces the information load by compressing visual tokens, and the Layer Dynamic Skipper, which mitigates the computational overhead by dynamically skipping unnecessary layers in the language model during both forward and backward passes. Our approach is broadly applicable to diverse MLLM architectures and has been extensively evaluated on multiple benchmarks, demonstrating its effectiveness and efficiency.

Recent multimodal large language models (MLLMs) widely adopt multi-layer visual feature fusion to enhance visual representation. However, existing approaches typically perform static concatenation or weighted aggregation after visual encoding, without intervening in the representation formation process itself. As a result, fine-grained details from early layers may be progressively suppressed during hierarchical abstraction. Moreover, directly introducing shallow-layer features into the language model often leads to semantic distribution mismatch with the visual feature space that the LLM's cross-attention layers were pretrained on, which typically requires additional adaptation or fine-tuning of the LLM. To address these limitations, we revisit visual representation learning from the perspective of representation evolution control and propose a cross-layer memory-modulated vision framework(SCVM). Specifically, we introduce a recursively updated cross-layer memory state inside the vision encoder to model long-range inter-layer dependencies. We further design a layer-wise feedback modulation mechanism that refreshes token representations at each layer based on the accumulated memory, thereby structurally regulating the representation evolution trajectory. In addition, we incorporate an auxiliary semantic alignment objective that explicitly supervises the final memory state, encouraging progressive compression and reinforcement of task-relevant information. Experimental results on multiple visual question answering and hallucination evaluation benchmarks demonstrate that SCVM achieves consistent performance improvements without expanding visual tokens, introducing additional vision encoders, or modifying or fine-tuning the language model.