ML: Attention

Machine-learning ideas tagged Attention in the ML taxonomy of the Math2NN corpus.

373 ideas found

Unverified 2026

Delocalization-regularized sparse masks

Use eigenvector delocalization as a mask-quality criterion rather than selecting a random sparse graph blindly. Penalize masks whose normalized adjacency has concentrated leading eigenvectors or disconnected or weakly connected components, while preserving the power-law distance prior. This creates a sparse routing graph that is less likely to trap information in local regions.

Useful5/10
Difficulty6/10
Novelty7/10
Paper: Emergent quantum chaos from correlations on a random graph arXiv:2607.11662
Unverified 2026

Ground-State Fractional Attention

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.

Useful5/10
Difficulty5/10
Novelty7/10
Paper: Sharp Broken-Power Lorentz Estimates for Fractional Powers of Radial Schrödinger Operators with Inverse-Square Asymptotics arXiv:2607.11280
Unverified 2026

Schatten Distance Fingerprint Regularizer

Represent tokens, features, or attention states by normalized rank-one matrices and train the network to preserve their Schatten-​p distance profiles over complex phase rotations. Because the paper proves that equality of all distances \(\|\lambda e-v\|_p\) identifies \({\rm Tr}(e^*v)\), this regularizer preserves matrix overlap geometry under a learned transformation.

Useful5/10
Difficulty6/10
Novelty8/10
Paper: Tingley's Problem for Schatten \(p\)-Classes, $0<p\ne 2<\infty$ arXiv:2607.11244
Unverified 2026

Dyck-Polytope Sparse Routing

Replace independent top-k expert or attention-edge selection with a gate vector constrained by hierarchical path budgets modeled on the paper's extended-Dyck-path polytope. Ordinary interactions receive continuous nonnegative capacities, while a designated class of cross-group interactions receives binary or clipped-to-one gates, producing structured sparsity and preventing many correlated paths from consuming the same routing budget.

Useful5/10
Difficulty6/10
Novelty7/10
Paper: FFLV bases for covariant representations of $\mathfrak{gl}(m|n)$ arXiv:2607.11133
Unverified 2026

Induced-Star-Free Stable Graph Propagation

Constrain a learned binary graph or sparse attention-routing graph so that every node neighborhood has no independent set of size k. This local anti-star condition gives an explicit upper bound on the graph Laplacian spectral radius, allowing a larger but certified stable diffusion step or residual propagation coefficient.

Useful5/10
Difficulty6/10
Novelty6/10
Paper: The largest Laplacian eigenvalue of induced-$K_{1,r}$-free graphs arXiv:2607.09390
Unverified 2026

Crossover-Aware Multiscale Attention

Build a neural architecture whose receptive field or attention span is increased according to an estimated disorder-to-order crossover scale. Local branches process windows below the crossover as if they were stochastic, while a global branch is activated only when the context exceeds the predicted scale needed to expose deterministic recurrence. This targets sequences or images containing long-range quasiperiodic, hierarchical, or algorithmically generated structure that is statistically…

Useful5/10
Difficulty5/10
Novelty7/10
Paper: The scales of disorder in perfect quasicrystals arXiv:2607.09274
Unverified 2026

Charge-conserving Z3 neural block

Represent each feature as belonging to one of three \(\mathbb{Z}_3\) charge sectors and constrain every linear and multiplicative operation to obey charge addition modulo 3. Add invariant cubic gates such as \(x_1x_2x_3\) or \(x_q^3\), which can express the same phase-insensitive interaction selected by the paper's three-photon drive. This should improve data efficiency and exact cyclic-augmentation consistency when the task has a genuine ternary symmetry.

Useful5/10
Difficulty4/10
Novelty6/10
Paper: Quantum and Classical Potts Criticality in Driven-Dissipative Bosonic Lattices arXiv:2607.08425
Unverified 2026

Lattice-Laplace Polytope Attention

Replace or augment conventional dot-product attention with features generated by a convex polytope's lattice Laplace partition function. For a query-dependent point inside a learnable polytope, the log-partition gradient is the expected lattice direction under a Gibbs distribution, while its Hessian is a covariance matrix that supplies curvature-aware features.

Useful5/10
Difficulty5/10
Novelty8/10
Paper: Dual Lattice Functions of Polytopes arXiv:2607.08101
Unverified 2026

Rayleigh–Jeans Condensation Routing

Replace softmax attention or dense MoE routing with a normalized Rayleigh–Jeans distribution over tokens or experts. If an item's energy is close to the chemical potential, its probability becomes disproportionately large, creating controllable low-energy condensation instead of the exponentially smooth allocation produced by softmax.

Useful5/10
Difficulty6/10
Novelty7/10
Paper: Thermodynamic description of worldwide distribution of energy and carbon emission arXiv:2607.07315
Unverified 2026

Masked Universal Host Layer

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.

Useful5/10
Difficulty5/10
Novelty5/10
Paper: A universal emulator for planar Ising lattices arXiv:2607.05308
Unverified 2026

Matroid-rank interaction bottleneck

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.

Useful5/10
Difficulty6/10
Novelty7/10
Paper: Analytic Spread via Linear Matroids arXiv:2607.07458
Unverified 2026

Four-Root Ternary Mixer

Replace a generic three-input concatenation MLP with a permutation-symmetric mixer built from the four signed combinations x+y-z, x-y+z, -x+y+z, and -x-y-z. Apply a shared truncated exponential to these combinations and aggregate symmetric pairwise products, producing controlled quadratic and higher-order interactions without materializing a full trilinear tensor.

Useful5/10
Difficulty4/10
Novelty7/10
Paper: On $η$-periodic Formal Ternary Laws arXiv:2607.06795
Unverified 2026

Random-label irregular sparse attention

Construct a sparse attention mask from a fixed regular candidate graph and one scalar random label per token, retaining edge $(u,v)$ when $x_u+x_v\geq\tau$. Unlike independent random pruning, this produces correlated neighborhoods and a controllable distribution of token degrees, potentially giving some tokens broad receptive fields while retaining a fixed sparse budget.

Useful5/10
Difficulty3/10
Novelty7/10
Paper: Irregular subgraph in a regular graph arXiv:2607.06465
Unverified 2026

Rigidity-Calibrated Set Attention

Augment pairwise attention on a set of n tokens with a rigidity operator derived from normalized pairwise directions. The operator couples infinitesimal node displacements through changes in pairwise distances, while the complete-graph theorem provides a geometry-independent eigenvalue target n/2 after spherical centering and normalization.

Useful5/10
Difficulty6/10
Novelty7/10
Paper: The Second Largest Eigenvalue of Stiffness Matrices of Normalized Complete Frameworks arXiv:2607.05472
Unverified 2026

Cyclotomic p-Cap Layer

Replace an ordinary token aggregation step with a p-replica cyclic-equivariant block. Features are copied into p replicas, processed by shared operators, coupled through a cap-like bilinear interaction, and projected onto cyclic invariants. An auxiliary commutation loss enforces that applying the operator before or after the p-fold lift gives similar outputs.

Useful5/10
Difficulty5/10
Novelty8/10
Paper: Noncommutative Cartier Formulae arXiv:2607.05360
Unverified 2026

Microscopic Boundary Pooling

Replace uniform set or point-cloud pooling with a microscopic weighting computed from pairwise feature-space distances. The resulting signed pooling vector should retain boundary and geometrically isolated points that ordinary mean pooling suppresses, potentially improving recognition when class information is concentrated on shape extremities or rare local configurations.

Useful5/10
Difficulty5/10
Novelty7/10
Paper: The microscopic weighting on a metric space arXiv:2607.05349
Unverified 2026

Quantile Envelope for Positive Kernel Layers

Add a differentiable rearrangement-envelope penalty to a positive integral-kernel layer. The penalty uses the Laplace-transform inequality to prevent the sorted upper tail of the layer output from becoming substantially larger than the cumulative upper tail of its input, providing a distribution-sensitive alternative to ordinary activation clipping or an L2 penalty.

Useful5/10
Difficulty4/10
Novelty7/10
Paper: On optimal endpoints for integral kernel operators arXiv:2607.05070
Unverified 2026

Piecewise-Symmetric Tensor Layer

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.

Useful5/10
Difficulty5/10
Novelty7/10
Paper: Piecewise Symmetric Tensors arXiv:2607.04712
Unverified 2026

Cofactor-Stable Attention

Treat each directed attention matrix as a graph transition matrix and form its Laplacian L = I - A. Compute the principal-cofactor vector to identify tokens with weak global access to the rest of the layer, and regularize the nonzero-eigenvalue product so attention does not become reducible or nearly singular. This targets pathological attention heads that isolate token groups and produce unstable or poorly propagated representations.

Useful5/10
Difficulty6/10
Novelty7/10
Paper: Voltage Stability Kernel: A Cofactor Theory of Voltage Stability in Lossy Power Systems arXiv:2607.02843
Unverified 2026

Bessel Totally-Positive Attention

Replace ordinary dot-product attention logits with a strictly totally positive kernel evaluated on positive, ordered scalar coordinates attached to queries and keys. Use the modified-Bessel kernel K(x,s)=I_s(x), whose every ordered minor is positive, then row-normalize it as an attention matrix. This creates an attention operator with a mathematically enforced anti-oscillatory structure rather than merely positive entries.

Useful5/10
Difficulty6/10
Novelty8/10
Paper: Strict Total Positivity from Spectral Darboux and Toeplitz Smoothing Mechanisms arXiv:2607.02778
Unverified 2026

Matroid-circuit equivariant message passing

Represent each matroid circuit as a structured hyperedge and perform message passing from circuit embeddings back to their constituent elements. Tie all circuit-update parameters that lie in the same automorphism orbit, so relabelings preserving the matroid produce exactly relabeled hidden states rather than requiring the network to learn this symmetry from data.

Useful5/10
Difficulty5/10
Novelty6/10
Paper: Encoding matroids into quantum states arXiv:2607.02736
Unverified 2026

Stochastic-order monotone attention ratios

Build an attention or positive-mixture module whose output ratio at two control settings is provably monotone in an ordered index such as token distance, retrieval rank, or discretized uncertainty. Use normalized-positive-series identities to replace an unstable quotient derivative with a difference of expectations, and penalize violations of the resulting stochastic-order condition during training.

Useful5/10
Difficulty5/10
Novelty5/10
Paper: A Probabilistic Sign Rule for Quotients of Positive Series and Integral Transforms arXiv:2607.02511
Unverified 2026

Log-Correlated Extreme-Value Logit Regularizer

Calibrate the maximum attention logit in each head against the log-correlated extreme-value law instead of applying fixed clipping or a fixed max-norm penalty. Penalize only maxima that exceed the predicted log N minus three-quarter log log N baseline by an unusually large order-one fluctuation, allowing ordinary sharp attention while suppressing rare pathological spikes.

Useful5/10
Difficulty5/10
Novelty7/10
Paper: Black Holes and Random Variables arXiv:2607.02233
Unverified 2026

Subcritical Preferential Sparse Attention

Construct a sparse attention or message-passing mask by sampling edges with preferential weights (d_u+alpha)(d_v+alpha), while keeping the edge count below the predicted connectivity threshold. This creates hub-like local communication patterns but prevents one giant component from forcing dense information mixing, reducing attention cost and potentially mitigating oversmoothing.

Useful5/10
Difficulty5/10
Novelty7/10
Paper: Sharp Asymptotics for the Largest Component in the Subcritical Regime of Preferential Attachment Without Vertex Growth arXiv:2607.00731