Matterhorn: Masked Time-to-First-Spike Encoding by Reassigning the Silent State for Sparse and Energy-Efficient Spiking Transformers

Zhanglu Yan, Kaiwen Tang, Zixuan Zhu +4

Spiking neural networks (SNNs) promise energy-efficient inference for large language models (LLMs), yet most reported savings rely on compute-operation counts that overlook data mo…