Skip to yearly menu bar Skip to main content


LyaNet: A Lyapunov Framework for Training Neural ODEs

Ivan Dario Jimenez Rodriguez · Aaron Ames · Yisong Yue

Hall E #310

Keywords: [ DL: Theory ] [ OPT: Control and Optimization ] [ DL: Algorithms ] [ Deep Learning ]


We propose a method for training ordinary differential equations by using a control-theoretic Lyapunov condition for stability. Our approach, called LyaNet, is based on a novel Lyapunov loss formulation that encourages the inference dynamics to converge quickly to the correct prediction. Theoretically, we show that minimizing Lyapunov loss guarantees exponential convergence to the correct solution and enables a novel robustness guarantee. We also provide practical algorithms, including one that avoids the cost of backpropagating through a solver or using the adjoint method. Relative to standard Neural ODE training, we empirically find that LyaNet can offer improved prediction performance, faster convergence of inference dynamics, and improved adversarial robustness. Our code is available at

Chat is not available.