Every idea extracted from recent arXiv mathematics papers — verified and unverified. Click an idea to open its full card; badges show the empirical verdict.
Add a finite-resolution geometric code to a 3D neural encoder: quantized lattice occupancy, local barycenters, and tangent directions are converted into structural tokens alongside ordinary point or mesh features. Choose lattice spacing from estimated local reach so that small perturbations do not change the code, and train the continuous encoder to agree with this discrete structural representation.
Represent sequence positions as vertices on a circle and use a maximal family of mutually non-crossing d-arcs to define the allowed attention interactions. Rotate the family by d positions, or use several phase-shifted families across successive layers, so each layer has only O(N) edges but repeated layers propagate information over long distances. The geometric compatibility rule replaces arbitrary local-window or hand-designed sparse masks with a structured family whose maximality gives broad…
Replace independent logits for all d-subsets with a neural head that outputs a d-by-n matrix A and assigns subset weight x_I=det(A_{:,I}). After normalization, these minors define a probability distribution over subsets. The head imposes a strong algebraic coupling between subset probabilities, reducing parameters and potentially improving extrapolation to rarely observed subsets.
Replace an unconstrained simplex router or differentiable mixture layer with a resource-cost-aware router whose learned costs satisfy the paper's monotonicity curvature condition. Use a Euclidean-regularized Frank–Wolfe oracle to update routing probabilities, which should reduce cycling and sensitivity when several examples or agents compete for the same experts.
Replace or augment relative-position attention with a positive fractional-integration mixing kernel whose radial behavior has separate inner and outer power laws. Tokens close to one another interact through the usual fractional singularity, while tokens near different radial scales receive a ground-state correction that can improve multiscale information transport without introducing a dense learned positional table.
Represent a neural network's categorical output over a rooted tree using cumulative probability mass on each rooted subtree. Train pairs of examples with a stochastic-dominance loss that compares these subtree masses, avoiding enumeration of all upper sets and making hierarchical monotonicity explicit. This is suitable for taxonomies, severity levels, hierarchical intents, and structured world-model states.
Estimate the simplex-based ratio of a target or learned convex piecewise-linear polytope and use the theorem \(\rho_\Delta(P)\le 2^d-1\) to choose a minimum useful ReLU depth. During training, monitor whether the learned polytope is approaching a high-\(\rho\) target; if it is, widen the model without increasing depth only when the diagnostic indicates that depth is the bottleneck.
Replace an unconstrained pairwise attention score with an intersection of coordinate-wise threshold or interval compatibility heads. Each head is a supergraph that permits pairs satisfying one constraint, while the final attention edge exists only when every head permits the pair. This provides an interpretable inductive bias for multi-constraint relations and prevents the model from approximating a conjunction using a single unstable nonlinear score.
Represent a population of N circular latent states using a three-parameter Möbius transformation applied to fixed uniform reference phases, rather than learning N unrelated angles. The resulting states remain on the circle by construction and can model concentrated or nearly uniform phase populations through a single concentration parameter.