Every idea extracted from recent arXiv mathematics papers — verified and unverified. Click an idea to open its full card; badges show the empirical verdict.
Represent uncertainty of a graph-structured neural feature field through dual covariance rather than explicitly storing a dense primal covariance. Recover calibrated primal marginal variances from dual statistics using the paper's covariance-complement identity.
Replace coordinatewise rounding of activation or embedding vectors with nearest-point quantization in a learned full-rank lattice. Learn an affine transform that makes the empirical activation region more isotropic, while regularizing the lattice covering density so it does not become inefficient as dimension grows.
Train a linear adapter between two representation spaces so that it preserves not only feature values but also the relative sparsity of sampled directions in the source representation subspace. Penalize the logarithmic spread between the largest and smallest support-size expansion ratios, preventing the adapter from making some directions dense while collapsing others. This is useful for transferring sparse features between checkpoints, aligning sparse autoencoders, or inserting a…
Replace an unrestricted collection of nested dyadic attention windows on a 2D token grid by a sparse antichain: no selected window may contain another selected window. Use the paper's exponential occupancy guarantee to control how many attention blocks reuse the same token, and add a differentiable log-moment penalty during training when exact antichain selection is relaxed. The expected benefit is bounded peak KV reuse and more predictable sparse-attention cost without discarding multiscale…
Replace a collection of dense task-specific linear layers with a common sparse structural matrix and task-specific edge strengths. All tasks share the same learned connectivity pattern, but retain independent values on active connections, allowing parameter sharing without forcing identical interactions.
Replace one-step greedy landmark selection in Nyström attention or kernel compression with a restricted pairwise-lookahead rule. The lookahead is motivated by the paper's explicit obstruction: a signed triangle can make individual column gains exhibit increasing rather than diminishing returns, so the best next column need not belong to the best pair.
Store quantized neural-network weights in ReRAM using GF(4)- or GF(8)-based constrained blocks rather than writing raw symbols. The encoder selects codewords whose local patterns cannot create the most damaging short rectangular sneak paths, while a decoder reconstructs the original quantized symbols after sensing. This targets persistent edge-model storage and memristor crossbar weight loading, where reducing read errors may be more valuable than the coding-rate loss.
Replace a dense or irregular binary interaction matrix in a value-aggregation layer with a signed sum of blocky masks. Each blocky mask groups a set of query rows with a disjoint set of key columns, allowing all queries in a group to reuse one summed value vector. This is most suitable for linear attention, graph message passing, or any layer where the interaction matrix is applied directly to values rather than passed through a row-wise softmax.
Add a structural positional channel formed from the Krylov sequence generated by the graph adjacency matrix and the all-ones vector. For graphs with k main eigenvalues, this sequence has rank k, so a GNN can retain all information obtainable from global walk counts using only k node features rather than storing many adjacency powers.
Add a bounded phase variable and a bank of local affine transport maps to an RNN or state-space model. The phase follows an irrational rotation, while the hidden state is transported through cells whose widths determine local gains, giving a controllable memory mechanism with analytically known distortion rather than an unconstrained recurrent Jacobian.
Build a deterministic complex-valued embedding for discrete IDs by evaluating finite-field polynomials through an additive character, but learn coefficients only for one representative of each Frobenius or cyclotomic orbit. The quotient removes parameters that generate exactly the same feature function after the trace map, avoiding flat optimization directions and reducing the size of the embedding layer.
Construct a sparse message-passing graph from a tree backbone by subdividing every backbone edge and attaching leaves so that 2d_T1(x_i)+f_i is constant across backbone vertices. Use this graph as a fixed communication skeleton, with propagation weights calibrated by the predicted spectral radius. The same construction can be compressed into an effective backbone operator by eliminating subdivision and leaf nodes.
Replace one-hot node IDs or large positional encodings in a GNN with coordinates from a compact abelian Cayley graph. The coordinates preserve graph-shortest-path geometry exactly, while Fourier characters of cyclic factors provide smooth neural features with fewer channels.
Represent many related sparse graph or attention patterns inside one fixed host connectivity pattern and activate each target instance with binary directional masks. The learned edge transformation and sparse-kernel layout are shared across instances, while the mask selects the target graph, enabling one compiled operator to process heterogeneous structures.
Construct a candidate feature for every edge pair or structured token pair, then retain a numerically independent subset under a feature-Jacobian matroid. The neural layer computes only the selected interactions, preserving directions that add new information rather than pruning solely by magnitude or attention score.
Build an implicit layer from a piecewise-linear maximal monotone operator on visible variables z_* and auxiliary variables z_**, then eliminate the auxiliary block rather than exposing it in the network output. Compute the layer through a fixed point of the eliminated component of a nonexpansive resolvent, with damping when the auxiliary map is not strictly contractive.
Use the paper's one-bit compressed-sensing lower bound to choose the number of binary latent measurements and to set a nonzero achievable-error floor during training. A sign bottleneck should not be given an unrealistically small bit budget: for approximately sparse latents, the target reconstruction error scales no faster than a power of effective sparsity divided by the number of sign measurements.
Add a topology-aware lower bound to point-cloud or graph token pruning: at each geometric scale, retain at least as many latent representatives as the persistent-homology rank between that scale and a larger scale. The method prevents the pruning module from collapsing independent connected components or cycles that remain persistent, while still allowing compression in topologically redundant regions.
Replace step-by-step hidden-state storage in a latent ODE, state-space model, or world model with a polynomial trajectory represented independently on short time blocks. At the end of each block, pass the next hidden state by summing temporal coefficients, allowing training and inference to discard the completed block while retaining a mathematically exact block interface.
Replace an unconstrained order-k weight tensor with a sum of components that are symmetric only within selected contiguous index blocks. This preserves interactions between blocks while tying parameters under within-block permutations, providing a tunable middle ground between a fully dense tensor and a fully symmetric tensor.
Replace an unconstrained linear residual adapter by an operator \(T\) satisfying a polynomial relation in the commutator operator \(\Delta_A(X)=AX-XA\). Choose the polynomial roots in a stable half-plane so that repeated commutators become nilpotent, making repeated adapter application terminate algebraically and permitting a finite-polynomial inverse of \(I+T\).