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 separate measurement-domain and target-domain token sets so a network can infer a field on one spatial domain from sparse observations on another in one forward pass. The same decoder can answer arbitrary target query points, avoiding an optimization loop for each inverse instance.
Use the activation singular-value spectrum to allocate a fixed zeroth-order parameter budget across layers instead of assigning the same rank everywhere. Layers with a large discarded singular-value tail receive more coefficient directions, while spectrally compressible layers use smaller adapters, preserving the gradient-relevant subspace under a global memory and query budget.
Add a fractional Laplacian penalty to neural functions over binary inputs so that high-order coordinate interactions are damped according to \(|S|^\alpha\), rather than treating all Fourier degrees equally. The penalty is estimated with random continuous-time bit-flip perturbations, requiring only extra forward passes and no explicit Fourier transform. It is especially suited to models that overfit through high-order Boolean interactions while retaining useful low-order structure.
Replace orthogonal Procrustes alignment between two latent dynamical systems with a learned bijection h that makes their transitions commute: h(f(z)) approximately equals g(h(z)). Parameterize h as an invertible affine map or coupling flow, allowing the correspondence to be non-orthogonal while retaining an exact inverse. The same constraint can be applied over multiple rollout steps, encouraging two models to represent the same computation even when their latent coordinates differ…
Train a neural controller as a uniformly accurate surrogate of a trusted but expensive controller, and use a measured small-gain condition to decide whether the surrogate is safe for closed-loop deployment. The approximation tolerance becomes an interpretable residual-state budget instead of an opaque validation metric.
Train a two-dimensional neural deformation map with the paper's Lp conformal-distortion energy instead of using only a determinant or smoothness penalty. The resulting barrier penalizes near-folds and directional collapse while permitting useful nonrigid deformation, making it suitable for spatial transformers, image registration, and learned coordinate warps.
Use the paper's inverse-temperature parameter to initialize networks containing m parallel depth-N branches. Choose branch count, depth, or an explicit aggregation scale so that beta = sqrt(2(N-1)/(n log m)) stays below the critical value sqrt(2), preventing the largest random branch from dominating the aggregate. This is applicable to residual multi-branch MLPs and other architectures whose block Jacobian is a sum of products.
Replace isotropic input-Jacobian regularization with a positive semidefinite, input-dependent metric learned jointly with the network. The metric uses diagonal scaling to suppress sensitivity in nuisance directions and a structured orthogonal rotation to discover combinations of input coordinates in which smoothness is task-useful.
Replace soft boundary penalties in neural operators with a hard projection onto a finite-dimensional span of homogeneous Dirichlet Laplacian eigenfunctions. Every projected hidden field is identically zero on the boundary, while increasing the number of retained eigenfunctions recovers the expressive capacity needed for operator approximation.
Replace magnitude-based channel or expert pruning with a subset-selection objective that maximizes the weakest direction in the candidates' activation span. Relax the binary mask to continuous gates, optimize an entropic soft minimum eigenvalue, and round the gates to retain a fixed number of channels or experts. This should preserve diverse representations and reduce redundant feature directions.
Use a minibatch-level transport plan that jointly decides which source graph should be paired with which target graph and how each target should be node-aligned. This can reduce total flow-matching displacement beyond per-example matching, producing shorter and less conflicting training trajectories without modifying the architecture.
Add a latent mode bank whose coordinates are learned by neural power iteration on observed state transitions rather than by jointly fitting an unconstrained latent dynamics model. Each mode is repeatedly regressed toward its one-step pushforward, normalized under the data distribution, and deflated against previously learned modes. The resulting latent coordinates are constrained to have approximately linear, diagonal dynamics, which should improve long-horizon prediction and make the…
Represent a predicted convex object by one point per prescribed unit direction and decode it as the convex hull of those points. Enforce direction-wise maximizer inequalities so every point is a genuine vertex, then use the covering-radius bound to choose the number and placement of directions according to the desired geometric accuracy.
Replace a large graph submodule by a compact boundary response operator that maps boundary node features to induced boundary fluxes after the interior has been eliminated. Stack these operators recursively to obtain a hierarchical graph neural network whose coarse-level computation preserves long-range effects of discarded vertices more faithfully than average pooling or simple node clustering.
Replace confidence-only masked diffusion decoding with an adaptive scheduler that chooses batches whose unrevealed tokens have low conditional total correlation given the already revealed context. The scheduler should preserve large parallel batches when token predictions are conditionally independent, but split highly dependent tokens into separate rounds to reduce forward-KL error.
Use the truncated Fourier representation of an irregular domain as a reusable spectral mask inside an FFT convolution layer. This gives a cheap alternative to point-cloud neighborhood aggregation while explicitly suppressing contributions from outside the physical domain and improving behavior near corners, cusps, and holes.
Train a conditional generator or set-valued predictor so that stochastic target refinements preserve the barycentric representation required by the source while allowing valid target-side diversity. The regularizer discourages collapse of multiple legitimate outcomes to one point without treating mean-preserving spread as semantic misalignment.
Construct a sparse radius graph over input samples or learned node embeddings, compute its cycle space, and remove the subspace generated by sufficiently short cycles. Feed the remaining quotient-cycle coordinates or Betti-rank estimate to a graph neural network as a global topological feature, or use them to guide pooling so that local redundant loops are collapsed while global loops are retained. The paper predicts that the threshold L approximately equal to |log r| graph hops is the critical…
Build a sequence or graph module whose state is a canonical set of affine tropical pieces rather than an opaque hidden vector. Compose consecutive segments by Minkowski addition of their lifted supports, merge alternatives by union followed by lower-hull reduction, and evaluate the resulting piecewise-linear function with a minimum.
Train a network to predict the context-dependent observation matrix rather than the latent inverse parameters themselves, then compute the latent parameters with a differentiable ridge-regression solve. This gives one model that can assimilate arbitrary observation vectors, exposes the conditioning of the inverse problem, and avoids forcing an MLP to learn the entire map from observations to parameters.
Replace uniform embedding dimensions with a globally budgeted allocation based on the estimated spectral complexity of each categorical feature. Tables whose category representations have large leading singular-value energy receive more dimensions, while high-cardinality tables are penalized because each extra dimension consumes more parameters.
Replace a Euclidean embedding bottleneck with a simplex-valued KL transport layer. The encoder maps each input to a positive probability vector, which is compared against learned positive stochastic prototypes through a c-convex log-sum-exp potential; the resulting barycentric or projected representation should suppress nuisance directions while retaining the topology of the data manifold.
Use an online estimate of the target distribution's remaining coordinate dependence to choose diffusion step sizes instead of using a uniform timestep grid. Take large jumps when the current state is nearly coordinate-factorized and shorten the step when the denoiser detects strong cross-coordinate dependence, targeting the paper's dependence-adaptive discretization scaling.
Replace independent top-k expert selection with a decision-diagram router that optimizes a quadratic surrogate over binary expert activations and continuous assignment weights. When expert redundancy or communication costs form a tree, partial routing decisions are merged whenever they have the same separator state, turning exponentially many candidate masks into a dynamic program over a small number of graph cuts.