IIT Guwahati Develops Brain-Inspired Energy-Efficient AI Model
| Category | Details |
|---|---|
| Institution | IIT Guwahati (SustainAI Lab, Mehta Family School of Data Science and AI) |
| AI Model Name | Spiking Heterogeneous Harmonic Resonate-and-Fire State Space Model (SH²RFSSM) |
| Inspiration | Mimics human brain neurons--activates only when processing critical information, reducing unnecessary energy use. |
| Key Technology | Combines spiking neural networks with state-space modelling to analyze long data sequences efficiently. |
| Neuronal Feature | Heterogeneous neurons--varied behavior improves real-world data pattern recognition. |
| Research Venue | Presented at International Conference on Machine Learning (ICML) 2026 in Seoul. |