Eventax enables exact-gradient training of arbitrary ODE-defined spiking neuron models in JAX via differentiable ODE solvers integrated with event detection.
Efficient recurrent architectures through activity sparsity and sparse back-propagation through time,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Training event-based neural networks with exact gradients via Differentiable ODE Solving in JAX
Eventax enables exact-gradient training of arbitrary ODE-defined spiking neuron models in JAX via differentiable ODE solvers integrated with event detection.