Resting Neurons, Active Insights: Improving Input Sparsification for Large Language Models
This work addresses efficiency and interpretability challenges in LLMs by improving input sparsification, though it is incremental as it builds on existing structural pruning techniques.
The paper tackled the performance gap in input sparsification for Large Language Models by introducing trainable spontaneous neurons to stabilize activations, resulting in substantial reduction of the gap while generalizing across tasks.
Large Language Models (LLMs) achieve state-of-the-art performance across a wide range of applications, but their massive scale poses significant challenges for both efficiency and interpretability. Structural pruning, which reduces model size by removing redundant computational units such as neurons, has been widely explored as a solution, and this study devotes to input sparsification, an increasingly popular technique that improves efficiency by selectively activating only a subset of entry values for each input. However, existing approaches focus primarily on computational savings, often overlooking the representational consequences of sparsification and leaving a noticeable performance gap compared to full models. In this work, we first reinterpret input sparsification as a form of dynamic structural pruning. Motivated by the spontaneous baseline firing rates observed in biological neurons, we introduce a small set of trainable spontaneous neurons that act as compensatory units to stabilize activations in sparsified LLMs. Experiments demonstrate that these auxiliary neurons substantially reduce the sparsification-induced performance gap while generalizing effectively across tasks.