Architecture ideas

Research ideas extracted from mathematics papers, categorized as Architecture.

Unverified 2026

Sparse Root-of-Unity Isometric Mixer

Replace a dense channel-mixing matrix by a sparse complex generalised weighing matrix W with exactly w nonzero entries in every row and column, then use U=W divided by square root of w as a norm-preserving mixer. Restricting to k=2 gives a real matrix with entries in {+1,-1}; k=4 supports signed phase rotations. The exact isometry should preserve signal and gradient norms while reducing channel-mixing cost.

Useful6/10
Difficulty6/10
Novelty7/10
Paper: Complex generalised weighing matrices in centraliser algebras of monomial representations arXiv:2607.16069
Unverified 2026

DynaBase Retrieval Forecast Head

Replace a parameter-heavy recurrent transition, or use this as a fallback, with a two-parameter nearest-neighbor successor blend in latent space. Given a query latent state, retrieve the closest state from an in-context trajectory and combine the query, the retrieved state, and its observed successor; this gives a zero-shot dynamical forecast with almost no trainable transition parameters.

Useful6/10
Difficulty4/10
Novelty5/10
Paper: A Minimal Interpretable Architecture for Zero-Shot Reconstruction of Dynamical Systems arXiv:2607.14937
Unverified 2026

Line-Graph Spectral Edge Parameterization

Replace one independently learned vector per graph edge with a truncated spectral expansion on the line graph. The model learns coefficients for low-frequency edge modes and reconstructs edge features before message passing, reducing parameters while imposing an inductive bias that incident edges should have correlated behavior.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: Lossy compression of weighted graph adjacency matrices by transform coding arXiv:2607.14834
Unverified 2026

Gaussian Simplex Classification Head

Replace the unconstrained final classifier with equal-norm regular-simplex class directions and train it under explicit isotropic Gaussian feature noise. At fixed signal energy and equal class priors, the paper's Gaussian-max theorem predicts that this geometry maximizes finite-noise maximum-likelihood decoding probability, making it a concrete candidate for robust classification heads.

Useful6/10
Difficulty4/10
Novelty4/10
Paper: Stochastic Domination of Gaussian Maxima: A Resolution of the Weak Simplex Conjecture arXiv:2607.14087
Unverified 2026

Hermite-Schatten spectral layer

Replace a dense learned linear operator on continuous or image features by a truncated Hermite projection expansion whose coefficients are directly regularized in a Schatten-p norm. The layer becomes a structured low-rank operator, while the radial Hermite-Laguerre correspondence provides an analytically tractable parameterization and an exact spectral penalty.

Useful6/10
Difficulty6/10
Novelty7/10
Paper: Quantitative Fourier Restriction Estimates for Weyl Operators: Fourier-Support Dependence and Lower Bounds arXiv:2607.13697
Unverified 2026

Finite-Orbit Circulant State Core

Replace the linear state transition in a small recurrent or state-space module by a circulant matrix acting on a vector over a finite field. The hidden state then has only finitely many possible values and follows an exactly periodic orbit after at most \(q^n\) states, eliminating numerical drift on modular-counting and symbolic-memory tasks. A learned real-valued encoder and decoder can surround the discrete core, while the transition itself is fixed, searched, or trained with a…

Useful6/10
Difficulty6/10
Novelty8/10
Paper: Periodicities in the Riordan arrays of polynomials over finite fields arXiv:2607.13442
Unverified 2026

Active-Set Constraint Projection Layer

Replace generic projected-gradient iterations for equality-plus-bilateral constraints with an active-set semismooth Newton layer. Each iteration fixes currently active lower and upper bounds and solves one structured saddle-point system, potentially converging in a few iterations when only a small subset of constraints is active.

Useful6/10
Difficulty6/10
Novelty6/10
Paper: Duality Framework for Flux Constrained Flow in Porous Media: Analysis and Numerics arXiv:2607.13256
Unverified 2026

SBP Energy-Stable Sequence Mixer

Replace a dense token-mixing matrix in a sequence model with a fixed or learnable SBP derivative operator D=P^{-1}Q. The discrete integration-by-parts identity makes the interior mixing energy-neutral or boundary-dissipative, reducing exploding activations in deep residual stacks while preserving directional information along the sequence.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: Gaussian FSBP operators: Comparison and application to numerical methods for hyperbolic conservation laws arXiv:2607.13224
Unverified 2026

Endpoint-transformed Hermite feature layer

Replace ordinary Fourier, polynomial, or raw-coordinate features for a bounded scalar coordinate with Hermite functions evaluated after a monotone endpoint transform. The transform sends endpoint singularities to localized tails on the real line, while a learnable scale controls how many Hermite modes are needed. This is suited to coordinate MLPs, neural operators, and implicit fields whose targets have square-root, logarithmic, boundary-layer, or derivative singularities.

Useful6/10
Difficulty4/10
Novelty7/10
Paper: Hermite spectral approximation for functions with endpoint singularities using exponential transforms arXiv:2607.12648
Unverified 2026

Hessian-guided orthogonal sparsification

Train a square orthogonal neural mixer while maximizing its entrywise fourth-power concentration. When optimization reaches a non-permutation stationary configuration, explicitly test rank-two row or column rotations and take a rotation with positive exact second variation, using the paper's constructive saddle-escape mechanism.

Useful6/10
Difficulty5/10
Novelty6/10
Paper: Local Maxima of the Entrywise $\ell_4$ Norm on the Orthogonal Group arXiv:2607.12431
Unverified 2026

Boundary-Only Cell-Complex Network

Parameterize a cell-complex neural network by features on p-cells and derive lower-dimensional boundary features using the cellular boundary map over F2. For a 2D square complex, neighboring plaquette bits determine each link feature through XOR, reproducing the paper's exact gauge-law reconstruction and preventing the network from representing inconsistent open boundary configurations.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: $p$-Form Gauge Dynamics and Digital Quantum Simulation -- Flux and Cosmological Constant Neutralization arXiv:2607.10950
Unverified 2026

Residual-Update Halting

Replace activation-magnitude-based adaptive computation halting with a criterion based on the actual recurrent update and a local stability margin. The loop halts when the state change is small relative to state scale for several consecutive steps, avoiding pathological decisions when LayerNorm-driven dynamics cause the activation norm to collapse.

Useful6/10
Difficulty5/10
Novelty6/10
Paper: LayerNorm as Implicit Gain Control in Looped Transformers arXiv:2607.10681
Unverified 2026

Young-Projected Tensor Channels

Decompose tensor-valued hidden states into invariant symmetric, alternating, and higher Young-symmetry channels before applying learned maps. This removes redundant tensor coordinates and prevents a neural layer from mixing incompatible representation types, yielding smaller equivariant modules with a cleaner inductive bias.

Useful6/10
Difficulty4/10
Novelty5/10
Paper: The Bernstein-Gelfand-Gelfand (BGG) Construction: Algebra, Geometry, and Analysis; Part I arXiv:2607.10662
Unverified 2026

Legendre-polynomial feature trunk

Construct a reusable ReLU trunk that emits approximate univariate powers or Legendre-polynomial features for each input coordinate, then combine them with a linear or low-rank polynomial head. This gives a compact explicit basis for smooth functions and can replace a large generic MLP in low-dimensional scientific regression or serve as a frozen or partially trainable front-end.

Useful6/10
Difficulty6/10
Novelty6/10
Paper: Approximation of Analytic Functions by ReLU Neural Networks with Adjustable Depth and Width arXiv:2607.10589
Unverified 2026

Exact simplex-lattice quantization

Replace independent coordinate rounding of a fixed-sum vector with nearest-point quantization in the projected integer lattice A_n^*. The quantized vector preserves the zero-sum constraint exactly, while the globally optimal rounding correction accounts for the aggregate residual induced by projection.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: A Faster Closest-Point Algorithm for the $A_n^*$ Lattices arXiv:2607.10479
Unverified 2026

Block-Probed Rational Spectral Layer

Replace a polynomial graph filter or repeated matrix multiplications in a graph neural network with a small rational filter evaluated at several shifts. Treat the incoming feature matrix as a block of probes rather than processing scalar probe vectors independently, allowing one set of shifted solves to expose multiple spectral directions simultaneously. The expected gain is higher approximation quality at the same number of operator applications, especially when the target filter has sharp or…

Useful6/10
Difficulty6/10
Novelty6/10
Paper: Convergence analysis of a nonlinear eigensolver based on rational approximation of the resolvent arXiv:2607.10377
Unverified 2026

Resolution-adaptive spectral front end

Replace a fixed Fourier or spectral resolution in a neural operator or sequence model with a data-adaptive spectral cutoff. Keep only modes whose estimated signal energy exceeds the noise-amplification and discretization floor implied by the available number of trajectories and samples per trajectory. This should reduce overfitting to high-frequency sensor noise and preserve accuracy when the same model is deployed at different sampling resolutions.

Useful6/10
Difficulty5/10
Novelty6/10
Paper: The Cost of Discretization in Functional Linear Regression: Minimax Rates and Adaptation arXiv:2607.09350
Unverified 2026

Local Characteristic Residual Gating

Transform local neural residuals into the Ripa model's characteristic coordinates before spatial aggregation, apply a mode-dependent gate based on neighboring characteristic jumps, and transform back. This lets the model damp oscillatory acoustic or equilibrium-mode corrections near discontinuities without globally smoothing every feature.

Useful6/10
Difficulty6/10
Novelty6/10
Paper: Fifth-Order Well-Balanced Path-Conservative A-WENO Scheme for the Ripa Model arXiv:2607.09293
Unverified 2026

Resolvent Fractional-Power Layer

Parameterize a learned feature-space operator as accretive but not necessarily symmetric, then apply its fractional power through a finite positive mixture of shifted resolvents. This provides a matrix-function layer that can represent directional and rotational interactions while avoiding unstable eigendecomposition of nonnormal matrices.

Useful6/10
Difficulty6/10
Novelty7/10
Paper: Functions and Means of Accretive Operators arXiv:2607.09152
Unverified 2026

Epoch-Frozen Masked Low-Rank Candidate Encoder

Add an online low-rank reconstruction layer before a recommender or ranking MLP. It estimates a latent action-feature subspace from masked candidate vectors, freezes that subspace for an epoch, reconstructs each partially observed candidate in the latent coordinates, and feeds only those coordinates to the predictor. The method is most promising when the ambient candidate dimension d is large but the effective rank m is small and missingness is not too severe.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: Stochastic Linear Bandits with Partially Observed Actions arXiv:2607.08971
Unverified 2026

Rank-Adaptive Tensor-Train MLP

Replace a large dense layer whose input and output dimensions factor into multiple modes by a TT-matrix whose parameters are stored as a chain of small cores. Periodically apply TT-SVD rounding to remove weak singular directions and keep the representation within a prescribed approximation error. This transfers the paper's central computational principle—perform tensor-product contractions directly in compressed form—to neural network layers.

Useful6/10
Difficulty5/10
Novelty4/10
Paper: A Tensor-Train Discontinuous Galerkin Method for the Vlasov-Maxwell System arXiv:2607.08936
Unverified 2026

Cross-Ratio Reversible Lattice Layer

Represent a hidden state as complex-valued points on a two-dimensional lattice and replace unconstrained local updates by the exact harmonic-quadrilateral completion rule from discrete conformal geometry. Given three corners of a plaquette, compute the fourth corner by a Mobius-rational formula enforcing cross-ratio minus one, then use a learned readout or forcing term for task-specific predictions. The layer supplies a hard geometric inductive bias and a directly measurable local constraint…

Useful6/10
Difficulty6/10
Novelty8/10
Paper: Integrability of Cauchy problems for discrete conformal maps and circle patterns arXiv:2607.08901
Unverified 2026

Task-Visible Axis Pruning

Prune hidden units only after testing whether their nonlinear gate is task-visible and downstream-used. For ReLU, a unit is removable when its preactivation does not cross zero on the task patch or its outgoing weight column is zero; this is a more structural criterion than weight magnitude and can be applied during width search or post-training compression.

Useful6/10
Difficulty3/10
Novelty5/10
Paper: Contravariance Theory: Strong Alignment for Minimal Solutions to Hard Tasks arXiv:2607.08561
Unverified 2026

Finite-group relative message passing

Use the quotient group's generator classes as a finite relation vocabulary and tie message functions by group displacement instead of by individual graph edges. This creates a compact, exactly consistent relation-aware GNN that can recognize repeated local structure and transfer parameters across graph instances sharing the same Cayley geometry.

Useful6/10
Difficulty5/10
Novelty8/10
Paper: Minimal Isometric Embeddings of Graphs into Cayley Graphs of Finite Abelian Groups arXiv:2607.07920