Every idea extracted from recent arXiv mathematics papers — verified and unverified. Click an idea to open its full card; badges show the empirical verdict.
Replace an unconstrained recurrent transition with a two-dimensional damped rotation whose parameters are induced by a learnable circular reorientation distribution. The first Fourier mode controls both memory persistence and phase rotation, giving the network an interpretable oscillatory memory while guaranteeing contraction when the effective decay rate is positive.
Replace a memoryless clipped recurrent output with a clipped observable plus a latent retained overshoot. The network exposes only a bounded output, but stores a fraction of the amount that would have exceeded the bound and feeds it into the next hidden-state update, allowing the model to represent persistent post-saturation effects without making the visible output unstable.
Replace magnitude pruning in a trained recurrent network with stochastic pruning probabilities computed from weight magnitudes and the covariance of neuron activities under injected noise. Connections whose endpoints fluctuate in a sign-compatible way receive higher retention probability, while retained weights are rescaled to preserve average recurrent strength. The method uses local weights and activity covariance, avoiding Hessian construction and expensive global saliency optimization.
Use the paper's finite-width O(n^{-1/2}) Gaussian-process approximation bound as a width-budgeting rule rather than choosing every hidden dimension uniformly. Estimate an architecture-specific constant for each layer or attention contraction, then allocate width according to the smallest dimension satisfying its allowed distributional error. This should produce narrower models at comparable GP-like behavior, or permit the same parameter budget to be concentrated in the layers where finite-width…
Initialize a stable diagonal state-space layer with decay rates \(\omega_i=|\xi_i|\), where \(\xi_i\sim\mathcal N(\mu,\sigma^2)\), instead of using a narrowly clustered rate distribution. The nonzero density of rates near zero creates a population of slow modes whose aggregate impulse response has an algebraic tail, enabling long-horizon memory while every finite-dimensional mode remains exponentially stable.
Add a bounded routing state to an RNN, state-space model, or mixture-of-experts layer, with several neutral fixed points representing persistent modes. The state moves between modes when far from a fixed point but escapes each mode only polynomially when close to it, creating controllable long memory without setting a linear eigenvalue arbitrarily close to one. A temperature parameter selects between an entropy-rich phase using many modes and a low-entropy phase concentrated near one preferred…
Replace an unrestricted additive recurrent or fast-weight memory with a sign-selectable update: for each incoming update vector, choose between adding and subtracting it so that a smooth compact potential of the memory state is minimized. This is appropriate when the memory representation has sign symmetry, such as signed random features or a learned linear sketch; it is not a drop-in replacement for ordinary gradient updates where the sign carries semantic information.
Replace unconstrained recurrent-state decay with a one-dimensional latent defect field whose states evolve by local diffusion and pair reactions. Defects can move over long distances and persist, while creation and removal occur only in pairs, giving the memory a structured cancellation mechanism that is potentially better suited to delayed-event and parity-like sequence dependencies than a standard GRU or diagonal SSM.
Parameterize a learned 3-state transition operator as a product of at most seven elementary row-stochastic matrices rather than learning its nine entries independently. Each factor performs one convex pull-in of row i toward row j, so every intermediate and final matrix remains row-stochastic and the layer has a sparse, bounded-depth interpretation.