AI RESEARCH

Closing the Theory-Practice Gap in Spiking Transformers via Effective Dimension

arXiv CS.AI

ArXi:2604.15769v1 Announce Type: cross Spiking transformers achieve competitive accuracy with conventional transformers while offering $38$-$57\times$ energy efficiency on neuromorphic hardware, yet no theoretical framework guides their design. This paper establishes the first comprehensive expressivity theory for spiking self-attention.