Every idea extracted from recent arXiv mathematics papers — verified and unverified. Click an idea to open its full card; badges show the empirical verdict.
Use one fewer point than the exact support threshold for a weighted multi-output linear head. The paper proves that the worst-case multiplicative loss penalty at this budget is only 1+1/(dm^2), giving a principled memory-saving option rather than an arbitrary subset-size heuristic.
Replace the recurrent transition or state-space mixer with a reversible transport followed by complementary relaxation of symmetric and antisymmetric feature components. The construction preserves a weighted energy and damps both parity sectors, giving bounded long-horizon powers without requiring the learned transition matrix itself to be symmetric. A numerical-range ellipse can be used as a cheap training-time certificate against transient growth.
Use exhaustive finite-state analysis as an architecture-search objective for coupled recurrent motifs. Instead of independently tuning a normalization gate and a WTA gate, enumerate their possible interfaces and select couplings that create group action only on a small joint image set, yielding controlled reversible subdynamics embedded in an otherwise dissipative system.
Constrain the learned coefficients of a high-order linear recurrent or state-space layer using the block Hurwitz matrix associated with its matrix characteristic polynomial. Penalize near-singular Hurwitz blocks and, for degrees two and three, optionally enforce positive leading Hurwitz determinants; use companion-matrix eigenvalues as the definitive stability check rather than trusting determinant positivity at degree four or above.
For a neural network predicting A coupled concentration or density fields, decompose Fourier-space fluctuations into a charge direction and its charge-neutral composition subspace. Hard-project the predicted fields to eliminate the global charge mode, and regularize their low-wavenumber covariance so that neutral modes retain finite susceptibility while the charge structure factor follows the Coulombic suppression S_ZZ(k) proportional to k squared. This should improve long-range physical…
Initialize or regularize recurrent matrices so that each unit receives an approximately cancelling sum of positive and negative weights, while keeping the global variance and spectral radius fixed. Sweep a continuous balance parameter instead of imposing balance blindly, because the paper predicts qualitatively different behavior for saturating, sub-linear, and odd nonlinearities.
Use the Legendre-dual Cramér transform of the minibatch embedding distribution as a geometry-aware novelty and confidence score. Add a penalty that separates correctly classified embeddings from high-Cramér-energy examples, or use the energy directly for OOD detection and hard-example mining; unlike squared distance, it adapts to anisotropic and non-Gaussian convex support.
Train a neural state-space model whose one-step dynamics are linear in a fixed analytic feature vector, and use the empirical feature Gram matrix to detect whether passive trajectories identify the dynamics. Add data collection or replay only when the Gram matrix is poorly conditioned; the analytic-feature assumption predicts that persistent excitation should emerge without deliberately visiting every operating mode.
Apply the paper's quiver-semistability viewpoint to a graph-structured state-space layer, treating each node's latent state space as a quiver vertex and each message-passing or coupling matrix as an arrow. Penalize approximately invariant collections of node subspaces that receive little signal from the input, so the learned latent dynamics cannot hide useful information in unreachable subnetworks. A dual output-side penalty can prevent predictive information from becoming confined to…
Replace an unconstrained recurrent transition matrix with a J-selfadjoint matrix A, where J is a fixed diagonal signature matrix with only a small number of negative entries. Add a sampled Kreiss-resolvent penalty to suppress transient amplification while preserving the expressive dimension of the hidden state. The paper's bound predicts that worst finite-time amplification depends on the smaller inertia index rather than the full hidden dimension.
Impose a screened pair-correlation dynamics on stochastic neural replicas so that correlation fluctuations relax locally instead of propagating across the entire representation. The key control knob is a learned or scheduled relaxation rate \(\mu_{FB}\), which predicts a measurable correlation length \(\xi_{FB}=\sqrt{D_{eff}/\mu_{FB}}\). This can be used as a locality regularizer for token representations, diffusion trajectories, or recurrent hidden states.
Represent a neural density or feature field by two positive reciprocal branches whose product is the modeled density, analogous to the forward and backward fields in the paper. Add stochastic branching perturbations to the two branches and train their cross-covariance so that the diagonal paired density matches the target while off-diagonal correlations remain finite-range. This creates a structured alternative to an unconstrained single-field uncertainty representation.
Replace the raw position/channel basis of a one-dimensional sequence module by eigenvectors of the projected cubic radial kernel matrix. Penalize or truncate coefficients in eigenmodes with many sign changes, giving a mathematically ordered smooth-to-oscillatory inductive bias while preserving the two-dimensional nullspace corresponding to affine trends.
Insert a differentiable reaction-diffusion layer that converts dense token or pixel features into sparse, spatially coherent routing masks. Two competing orientations form complexes through conserved monomer reservoirs, so local assignments can cluster while opposite assignments mutually exclude one another instead of independently activating at the same location. The layer can be used as a soft-to-hard MoE router, attention-mask generator, or object-part grouping module.
Construct a graph-based latent state whose velocities evolve through free-flight updates and pairwise elastic collision operators. Each collision operator is orthogonal, so total latent kinetic energy is exactly conserved; a connected interaction graph is intended to eliminate unwanted component-wise polynomial invariants and improve long-horizon stability.
Replace ordinary parameter updates for a constrained neural network with an annealed reciprocal-manifold flow. Each differentiable inequality constraint remains strictly satisfied during the optimization trajectory, avoiding projection or a per-step quadratic program. This is most useful for safety-critical policy learning, bounded network outputs, parameter-budget constraints, or training with explicit robustness inequalities.
Regularize a neural representation so that no one-dimensional projection places too much probability mass inside a narrow interval. This transfers the paper's uniform tube estimate into an anti-collapse constraint, making representations robust to adversarial directions and preventing hidden features from becoming effectively low-dimensional.
Use cumulative control measures as the input to a neural surrogate instead of the full sequence of control values. For a quasi-static free-boundary system satisfying the paper's average-pressure path-independence assumption, two nonnegative control histories with identical integrals up to time t should produce the same state at t, allowing a smaller training input and fewer distinct control trajectories in the dataset.
Construct a neural state-space model with an explicit first-order input-to-output path instead of forcing every output to depend only on deeply propagated hidden states. Penalize or reject learned linearizations whose transfer matrix has relative degree greater than one, then train a storage-function certificate for the remaining passive dynamics. This preserves the paper's relative-degree compatibility condition while allowing high-order internal memory.
Use two points with approximately equal minibatch loss to construct an ellipcenter: the intersection of the normal lines through the two points, where the normals are their gradients. The resulting update uses local curvature information in the span of two gradients and can be relaxed toward the current parameters or combined with momentum.
Represent binary nuisance variables or augmentation bits as coordinates of a Hamming cube and penalize the model response that remains predictable from a random k-coordinate subset. Use the theorem's derivative-plus-global-norm certificate as the regularizer, retaining its p/log p dependence instead of using an arbitrary masking penalty. Random coordinate permutations provide a cheap stochastic approximation to the subset average.
Build actuator or parameter constraints directly into the neural controller using a differentiable hard-saturation map rather than penalizing violations after the fact. This makes the Lyapunov certificate apply to the actual bounded controller and prevents training from exploiting unrealistically large actions.
Add a differentiable or inference-time projection to mesh and graph neural operators that contracts each predicted nodal state toward a weighted cell anchor. The anchor is the geometry-weighted mean, so the correction preserves the weighted integral exactly, while the contraction parameter is chosen to keep all nodal states inside a convex physical set such as positive density and energy or a probability simplex.
Use the OU process driven by multiple dichotomous noises as a bounded colored-noise module for latent-variable or diffusion sampling. Its stationary forcing is compactly supported for fixed amplitudes, while heterogeneous amplitudes and switching rates create controllable non-Gaussian structure before the large-K Gaussian limit.