A neuromodulable current-mode silicon neuron for robust and adaptive neuromorphic systems
The paper presents a novel current-mode silicon neuron that incorporates neuromodulation to achieve robust and adaptive computation, a feature largely unexplored in neuromorphic circuits. A mathematical model and tuning tools are introduced, and both theoretical analysis and experimental results on a low‑power 180 nm CMOS implementation confirm the neuron’s biologically plausible adaptation across a wide range of parameters. The design demonstrates high robustness, flexibility, and scalability over varying currents and temperatures, positioning it as a strong candidate for real‑world neuromor…
The paper presents a novel current-mode silicon neuron that incorporates neuromodulation to achieve robust and adaptive computation, a feature largely unexplored in neuromorphic circuits. A mathematical model and tuning tools are introduced, and both theoretical analysis and experimental results on a low‑power 180 nm CMOS implementation confirm the neuron’s biologically plausible adaptation across a wide range of parameters. The design demonstrates high robustness, flexibility, and scalability over varying currents and temperatures, positioning it as a strong candidate for real‑world neuromor…
Swiss AI research emphasizes energy‑efficient, robust neuromorphic hardware, and this neuron’s low‑power CMOS implementation and adaptive capabilities directly support those objectives.
Original source record. Open the original record to verify the underlying announcement.