news · 2 min
THOR x NeuroBench Challenge 2026: Phase-I Results Announced
Celebrating groundbreaking accuracy and energy efficiency decoders on open BCI benchmarks
The Neuromorphic Commons (THOR), in partnership with NeuroBench and the Edge AI Foundation, is proud to announce the final evaluation results for the inaugural THOR x NeuroBench Challenge 2026 (Phase-I).
Bringing together international researchers and spike-based algorithmic pioneers, Phase-I tasked participants with decoding continuous motor imagery EEG signals (OpenBMI dataset) across two rigorous evaluation tracks: absolute classification accuracy and compute/energy efficiency measured by synaptic operations (SynOps) and memory footprint.
Finalist teams demonstrated that biologically inspired spiking architectures can achieve competitive classification performance while significantly reducing computational overhead compared to conventional ANNs.
Explore the full finalist breakdown, consensus findings, Pareto frontier charts, and award winners on the THOR x NeuroBench Challenge Leaderboard.
