Unverified
2026
Generate memory slots or attention keys by applying learned unitary transformations to one seed vector instead of storing every slot independently. Select transformations sequentially when they add a sufficiently new direction and reject phase-equivalent or highly coherent candidates, targeting a well-conditioned near-Parseval frame.
Useful6/10
Difficulty6/10
Novelty7/10
Unverified
2026
Replace or augment a dense attention similarity matrix with a Min-cone matrix generated by a monotone scalar sequence. The resulting matrix is positive semidefinite by construction, has only O(n) learned scalar parameters, and can be multiplied by values in O(n d) time using cumulative sums rather than forming an n-by-n matrix.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Insert a fixed or learnable multiresolution transform before a CNN or vision-transformer block and penalize its coefficients with the paper's local tent-space square function. The penalty couples coefficients belonging to the same spatial dyadic region and can remove localized multiscale feature packets, potentially producing structured sparsity and better denoising than independent l_1 shrinkage.
Useful6/10
Difficulty4/10
Novelty6/10
Unverified
2026
Replace the length-L learned convolution kernel in a causal sequence layer with K Laguerre basis functions, where K is much smaller than L and the basis parameter controls the decay time scale. The layer retains a long receptive field but learns only K coefficients, while FFT or a fixed state-space realization evaluates the resulting convolution efficiently. This is especially appropriate for audio, sensor streams, and long-context regression where the desired impulse response is smooth or…
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Pack two real tensors into the real and imaginary components of one complex tensor, perform one complex transform, and recover both real spectra through conjugate symmetry. Combine this with implicit padding so separate padded real arrays and transform workspaces are not allocated.
Useful6/10
Difficulty6/10
Novelty5/10
Unverified
2026
Represent a 3D neural field using high-order polynomial coefficients attached to an adaptively refined tetrahedral mesh, with a small MLP predicting residual corrections from local coordinates. Refine only tetrahedra whose prediction, rendering, or PDE residual is large, and use globally ordered vertices so neighboring tetrahedra share identical face and edge coefficients without hanging-node constraint solves.
Useful6/10
Difficulty6/10
Novelty7/10
Unverified
2026
Use the number of tokens participating in an attention operation to adapt the projected query/key dimension according to the sharp Johnson–Lindenstrauss bound. Short attention windows receive a dimension limited by n-1, while longer windows use a logarithmic dimension, avoiding a fixed large head dimension when the current attention set contains too few points to require it.
Useful6/10
Difficulty6/10
Novelty6/10
Unverified
2026
Separate low-support monomials, which involve only a few distinct input coordinates, from high-support monomials in a high-degree symmetric interaction layer. Compute the low-support orbit features exactly and prune, sample, or factorize the high-support tail, using the paper's cutoff scale as the initial sparsity rule.
Useful6/10
Difficulty5/10
Novelty8/10
Unverified
2026
Insert a hyperspherical adapter that splits an embedding into several unit-sphere blocks, changes the dimension of each block, and recombines them with a synchronized spherical join. Train the adapter to preserve pairwise angular distances, while using the paper's max-distortion composition principle to avoid uncontrolled accumulation of blockwise errors.
Useful6/10
Difficulty5/10
Novelty7/10
Unverified
2026
Place doubly stochastic stream mixing immediately before a quantizer, activation compressor, or latent bottleneck and jointly optimize task loss with estimated code length. The paper's entropy argument says that this linear mixing cannot increase differential entropy, so it can provide cross-stream representation capacity without an ideal entropy-rate penalty; the entropy bottleneck then learns which feature values deserve bits.
Useful6/10
Difficulty4/10
Novelty4/10
Unverified
2026
Regularize neural features indexed by a compact transformation group using the three conditions from the vector-valued Pego theorem: nearby group transformations should produce nearby features, high group-Fourier coefficients should have small energy, and feature energy should remain concentrated in a fixed low-dimensional value-space subspace. The third term is important for large or effectively infinite-dimensional feature spaces, because translation and Fourier smoothness alone do not…
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Apply the paper's weighted nuclear elastic-net principle to the transition matrix of a recurrent or linear state-space layer. Penalize low-rank structure after whitening by the observed hidden-state covariance, while retaining a ridge term that prevents poorly excited state directions from producing unstable or arbitrarily large transition weights.
Useful6/10
Difficulty6/10
Novelty5/10
Unverified
2026
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…
Useful6/10
Difficulty5/10
Novelty8/10
Unverified
2026
Replace random or k-means initialization of a k-expert router with a moment-based range finder on a calibration batch of hidden states. Estimate a low-dimensional second-moment subspace, enlarge it using one-free-index third-Hermite contractions, and fit the router's expert centroids and weights only in this resulting subspace. The router can then operate on projected hidden states while retaining an optional small residual adapter.
Useful6/10
Difficulty5/10
Novelty7/10
Unverified
2026
Insert a proximal layer after a graph, mesh, or spherical convolution that groups all coordinates belonging to the same Laplacian eigenspace and applies one shared shrinkage gate to the whole group. Unlike coefficientwise spectral pruning, the result is unchanged if the eigenvectors inside a repeated eigenspace are rotated, preventing arbitrary basis-dependent feature selection.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Use the paper's explicit pixel-spacing error bound to make radial topological features and losses resolution-aware. Treat intervals whose endpoint changes are below the discretization tolerance as unreliable, and use the bound to select contour resolution or a persistence threshold instead of tuning these quantities arbitrarily. This can improve robustness to rasterization, small contour perturbations, and multi-resolution training.
Useful6/10
Difficulty4/10
Novelty7/10
Unverified
2026
Make Laplacian latency attention hardware-friendly by quantizing the exponential decay and row normalization to powers of two. Exponentials and divisions can then be implemented with lookup tables and integer shifts, enabling a low-bit attention kernel for neuromorphic and edge accelerators.
Useful6/10
Difficulty4/10
Novelty7/10
Unverified
2026
Approximate a long hidden-state or key/value trajectory by a small sum of exponential modes, storing only the mode nodes and vector weights instead of every timestep. Recover the modes from a Hankel matrix through an annihilating polynomial, then reconstruct the trajectory or a compact recurrent state during inference.
Useful6/10
Difficulty6/10
Novelty5/10
Unverified
2026
Give a token, patch, retrieval-item, or expert-selection module a learned utility f_theta(S) over subsets S, and penalize violations of the paper's submodularity and strong-submodularity inequalities. The resulting selector should prefer complementary elements: the marginal value of adding an item decreases when the current selected set is already rich in similar information. At inference, use greedy marginal-gain selection rather than independently thresholding token scores.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Add a neural feature, adapter, or expert block only through the component of its outputs that is orthogonal to the span of all previously installed blocks. Quotient coefficient directions that produce nearly identical outputs with an SVD or pseudoinverse, so the new block contributes intrinsic representational dimensions instead of duplicating old features. The expected benefit is a smaller effective architecture and better-conditioned block expansion.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Replace an unconstrained recurrent transition with a second-order resonant state whose restoring matrix is full-rank but whose damping is low-rank. The low-rank damping creates a small set of rapidly controlled bright modes and a large dark subspace with long memory, while a small optional damping term prevents numerical drift in completely dark modes.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Construct a sparse recurrent network with positive edge weights and Leaky-ReLU updates so that one selected hidden node, observed over a finite time window, contains enough information to reconstruct the full hidden state. Add an auxiliary decoder from the observed trajectory to the initial state or current state, and use graph rewiring or edge-growth until every hidden node has a directed path to the sensor within the observation horizon.
Useful6/10
Difficulty5/10
Novelty8/10
Unverified
2026
Replace isotropic variance control in a bottleneck or router with a spectral polarization penalty that drives each latent direction toward either variance 0 or variance 1. The intended result is an automatically selected active subspace: inactive coordinates can be pruned or quantized aggressively, while active coordinates retain information instead of being uniformly attenuated.
Useful6/10
Difficulty4/10
Novelty7/10
Unverified
2026
Replace a learned dense bilinear map with a structured family of signed orthogonal matrices. Given feature vectors y,z in R^n, produce r interaction features h_a = y^T H_a z / sqrt(n), where the H_a form a Hadamard/Clifford-like family; the resulting bilinear map has operator norm at most one when r is within the Hurwitz–Radon limit. Learn only channel projections and optional scalar gates around this fixed mixer.
Useful6/10
Difficulty6/10
Novelty7/10