Architecture ideas

Research ideas extracted from mathematics papers, categorized as Architecture.

Unverified 2026

Curved-Critical Residual Lattice

Construct a 2D recurrent or residual neural lattice with slowly varying local couplings, while parameterizing those couplings so that an anisotropy invariant remains constant across all spatial and depth locations. The network obtains controlled local propagation velocities rather than arbitrary inhomogeneous amplification, enabling depth-dependent receptive fields while preserving near-critical signal propagation.

Useful6/10
Difficulty6/10
Novelty8/10
Paper: Two dimensional inhomogeneous classical systems at criticality arXiv:2608.02903
Unverified 2026

Diamond-Consistent Two-Route Layer

Construct a neural layer with two independently ordered transformations and train its operators to satisfy the paper's diamond equations, so that applying direction 1 then direction 2 gives the same result as direction 2 then direction 1. Unlike ordinary weight sharing, the mixed identity permits noncommuting operators whose interaction defects cancel exactly.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: Non-Abelian Hirota-Miwa Equations for the KPZ Universality Class arXiv:2608.02772
Unverified 2026

Two-chart refinement cell

Build a recurrent ReLU module that evaluates a binary refinement cascade using a fixed-dimensional state and shared cell weights. Replace hard binary digit selection with two overlapping circle coordinates; switch between their affine state updates only at points where the two candidate updates agree, so the switch is an exact continuous piecewise-linear ReLU operation rather than a multiplicative gate.

Useful6/10
Difficulty6/10
Novelty7/10
Paper: Neural network realization of binary refinement iterates via a two-chart atlas selector arXiv:2608.02624
Unverified 2026

Tangential FFN Residuals

Project each FFN residual update onto the tangent space of the current token residual direction before adding it to the stream. This preserves the component that changes representation direction while suppressing norm-only motion, which may reduce residual-norm drift and aggregation-induced representation collapse.

Useful6/10
Difficulty4/10
Novelty6/10
Paper: Feed-Forward Steering in Transformer Residual Dynamics arXiv:2608.02071
Unverified 2026

Noncommutative controllability regularizer

Equip a recurrent or state-space layer with multiple noncommuting transition operators and regularize the span of finite operator words applied to the input injection matrix. This discourages hidden directions that cannot be reached from the input and may improve long-range input influence, gradient propagation, and robustness under operator switching.

Useful6/10
Difficulty5/10
Novelty8/10
Paper: A noncommunicative Kalman condition for null controllability of backward stochastic parabolic systems arXiv:2608.01836
Unverified 2026

Fixed/Anti-Fixed Signed Message Passing

Replace a conventional signed-graph message-passing layer with two coupled feature channels: a fixed channel invariant under switching and an anti-fixed channel that changes sign under switching. Unsigned aggregation updates invariant features, while signed aggregation updates anti-invariant features, implementing the paper's sphere-plus-involution representation at the hidden-state level.

Useful6/10
Difficulty4/10
Novelty5/10
Paper: A Lovász Theta Parameter and Theta Body for Signed Graphs arXiv:2608.01716
Unverified 2026

Random-start learned AMP

Construct an unrolled phase-retrieval network that begins with an isotropic Gaussian estimate rather than a spectral initializer. Retain the AMP residual correction and Onsager subtraction, but learn the scalar measurement denoisers and step sizes; use several random starts and select the iterate with the lowest measurement residual.

Useful6/10
Difficulty5/10
Novelty4/10
Paper: Approximate Message Passing with Random Initialization for Phase Retrieval arXiv:2608.01654
Unverified 2026

Exact Involutive Sprugnoli Mixer

Replace a learned sequence-mixing matrix with a structured lower-triangular Sprugnoli operator whose square is exactly the identity. Applying the same operator in reverse reconstructs activations exactly, so it can be used as a reversible Transformer mixer or reversible channel permutation while retaining nontrivial long-range mixing.

Useful6/10
Difficulty6/10
Novelty7/10
Paper: Square roots in the Appell group and Sprugnoli arrays arXiv:2608.01497
Unverified 2026

U-centered relational attention

Replace raw pairwise attention or graph-edge scores by exact U-centered residuals, removing additive effects attributable to either endpoint. The resulting scores represent interaction beyond independent source and destination biases and satisfy zero row sums, preventing a few high-degree or high-activation tokens from dominating relational aggregation.

Useful6/10
Difficulty3/10
Novelty6/10
Paper: U-centering as subset ANOVA: edge regression and higher-order theory arXiv:2608.01364
Unverified 2026

Summed Resolvent Spectral Layer

Replace a recursive product implementation of a rational spectral filter with an additive sum of independently evaluated resolvents. Use the layer on a graph Laplacian, token-similarity operator, or other sparse feature operator to obtain a high-order filter without multiplicative roundoff and gradient amplification; the independent solves can also be batched or distributed across devices.

Useful6/10
Difficulty6/10
Novelty6/10
Paper: Stable and Efficient One-Way Modelling of Convective Disturbances in Laminar Boundary Layers: OWNS-Summation arXiv:2608.01155
Unverified 2026

Folded cluster feature mixer

Build a neural feature-mixing block from a sparse exchange matrix and a fixed sequence of cluster-style mutations. Transpose-related channels share parameters through orbit tying, while positive multiplicative-additive updates are evaluated with log-sum-exp. The block is a concrete alternative to a dense MLP projection and can be tested at equal parameter count and compute.

Useful6/10
Difficulty5/10
Novelty8/10
Paper: Cluster Algebras for Bosonic Plethysm arXiv:2608.00963
Unverified 2026

Outlier-Spectral Landmark Attention

Construct a symmetric token affinity matrix and approximate only its spectrally outlying token-mixing modes using a small set of sampled landmark columns. The resulting low-rank operator replaces an \(O(n^2)\) dense mixer by two skinny matrix multiplications, while the paper's residual guarantee predicts that large-magnitude global interaction modes are preserved.

Useful6/10
Difficulty5/10
Novelty5/10
Paper: Sublinear Time Eigenvector Approximation via Column Sampling arXiv:2608.00840
Unverified 2026

Schur-Agler contractive gated layer

Replace an unconstrained recurrent or residual transition with a block-structured contractive realization whose input-dependent multiplier is a direct sum of bounded branches. The resulting layer has a built-in non-expansive energy bound on the chosen normalized input domain, while still allowing different branches to respond to different coordinates or feature groups.

Useful6/10
Difficulty5/10
Novelty5/10
Paper: Function theory of the hexablock and applications to the tetrablock and Euclidean biball arXiv:2608.00819
Unverified 2026

Positive Bidiagonal Diffusion Layer

Replace a learned nonnegative banded transition matrix by an ordered product of learnable stochastic bidiagonal factors. Each factor performs one local left or right transport step, so a product of p lower and q upper factors creates an effective bandwidth of p+q while retaining nonnegative entries, row sums equal to one, and a highly structured propagation kernel.

Useful6/10
Difficulty3/10
Novelty7/10
Paper: Positive Bidiagonal Factorizations for Banded Markov Processes arXiv:2608.00788
Unverified 2026

Prototype Distance-Field Safety Layer

Store a finite library of successful robot configurations or action-conditioned waypoints and construct a smooth soft minimum of their distances. Use the negative distance gradient as a structured action prior, add a learned residual policy, and pass the combined action through a quadratic-program safety layer. This gives a neural controller an explicit attraction basin toward demonstrated solutions while preventing violations of known state constraints.

Useful6/10
Difficulty5/10
Novelty6/10
Paper: Grasp Execution Without a Planner: Configuration-Space Grasp Distance Fields with Certified Safety & Guaranteed Quality arXiv:2608.00600
Unverified 2026

Polylogarithmic Heavy-Tail Attention

Replace dense attention or strictly local windows with a refreshed sparse communication graph whose edge lengths follow a truncated power law. For sequence positions or image patches at distance $r$, sample edges with probability proportional to $(1+r)^{-(d+\alpha)}$; in the regime $0<\alpha<d$, repeated layers should spread information across the sequence much faster than a local window while using only $O(nk)$ edges per layer. The main test is whether this preserves long-range accuracy at…

Useful6/10
Difficulty5/10
Novelty6/10
Paper: The Heavy-tailed Frog Model arXiv:2608.00399
Unverified 2026

Contracting Countable-Branch Router

Construct a routed neural state update from a collection of branch maps whose inverse-style refinement operators are uniformly contractive. The discrete routing sequence acts as an expandable symbolic code, while contraction makes the continuous state associated with a long routing history insensitive to initialization and earlier perturbations. Use a finite active top-\(K\) set during training, but retain an expandable branch table so the model can represent increasingly complex or rare modes.

Useful6/10
Difficulty6/10
Novelty7/10
Paper: Infinite-Piecewise Expanding Maps: Chaos, Ergodicity and Invariant-Set Complexity arXiv:2608.00398
Unverified 2026

Invariant-Preserving Latent Compression

Replace unconstrained low-rank compression of a neural state with an augmented basis that always contains vectors representing known conserved quantities or diagnostically important linear statistics. After each learned transition, project the state back onto the affine constraint set with an exact minimum-norm correction, preventing rank truncation and model error from accumulating in those statistics.

Useful6/10
Difficulty5/10
Novelty6/10
Paper: Structure-Preserving Dynamical Low-Rank Approximations for Stochastic Vlasov--Poisson Equations arXiv:2608.00397
Unverified 2026

Sparse Learnable Power-Law Head

Attach a symbolic sparse head to a neural encoder instead of using a dense final MLP. The head evaluates a library of learnable power-law and interaction terms on nonnegative learned features, jointly optimizes linear coefficients and exponents, and removes inactive terms with coefficient sparsity. This should provide a compact model with better relative-error behavior on positive targets spanning several orders of magnitude.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: Discovering Explicit Magnetic Core Loss Equations via Learnable Symbolic Sparse Identification arXiv:2608.00379
Unverified 2026

Spectator-canceling curvature router

Replace or augment a mixture-of-experts router with a relative transverse-curvature score computed between experts, rather than relying only on the router MLP logits. Experts that provide a broader, less stiff local response in task-relevant directions receive higher routing probability, while common nuisance or spectator directions cancel from the comparison. The score is invariant under a common linear reparameterization of the routing coordinates and can be restricted to a low-dimensional…

Useful6/10
Difficulty6/10
Novelty7/10
Paper: Channel selection at identically vanishing dissipation difference: isolating the frenetic sector of the overdamped path measure arXiv:2608.00041
Unverified 2026

Exact-Distance Landmark Attention

Replace dense spatial attention on a 2D feature grid by attention from every non-landmark token to a small periodic set of landmark tokens. Choose landmarks using the paper's residue construction so that every non-landmark token has at least one landmark exactly k Manhattan steps away. Use this guaranteed edge as a sparse global-information route, combined with a cheap local convolution or residual path for accuracy.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: Exact-Distance Domination in Grid Graphs arXiv:2607.29648
Unverified 2026

Finite-Band Polynomial Seed Bank

Construct one reference Lanczos basis for a symmetric propagation operator H, then derive several seed-specific spectral responses for Q_a(H)x_0 through finite-band polynomial connectors. With degree-r seeds, each transformed basis vector uses at most 2r+1 neighboring reference basis vectors, avoiding a separate Lanczos factorization for every seed.

Useful6/10
Difficulty6/10
Novelty7/10
Paper: Polynomial Initial-State Jumps and Christoffel Transforms in Krylov Complexity arXiv:2607.05294
Unverified 2026

Competing Infection-Removal Graph Layer

Replace a conventional graph message-passing layer with a finite-horizon stochastic propagation process containing susceptible, infected, and removed feature states. Messages spread along active infected-to-susceptible edges, while infected nodes are simultaneously deleted at a rate proportional to their susceptible-neighbor count. This provides explicit propagation control and anti-oversmoothing dynamics instead of repeatedly averaging over every neighbor.

Useful6/10
Difficulty5/10
Novelty6/10
Paper: The Zombie Infection Model arXiv:2607.29409
Unverified 2026

Signed-base fractal positional features

Augment standard Transformer positional embeddings with coordinates generated by the paper's signed-base digit expansion. Previous binary digits determine the sign and scale of later contributions, while a two-state Markov chain controls correlations between digits. This supplies multiscale positional structure using a small number of transition and base parameters.

Useful6/10
Difficulty4/10
Novelty7/10
Paper: Fractal random variables defined by probability distributions of digits of their $G_2$-representation having two bases with different signs arXiv:2607.29327