MinMax RNCs Match Regular Languages With No Vanishing Gradients
June 9, 2026
MinMax Recurrent Neural Cascades use MinMax algebra recurrence to achieve formal expressivity over regular languages, stable bounded activations, and non-vanishing gradients across all sequence lengths. They support parallel-scan evaluation at logarithmic depth and linear work, addressing core RNN training pathologies without sacrificing expressivity.
HOW THIS AFFECTS YOU
●
researcherWorth watching because it offers a theoretically grounded alternative to standard RNNs with provable gradient bounds and parallel evaluation — potentially relevant to sequence modeling architecture comparisons.