Attention variants, state-space and recurrent cells, normalization and token-mixing schemes — each tested against the standard block it replaces.
Replace dense token attention with K asynchronous local nonlinear consensus updates on a sparse token graph. Each selected token is moved to the unique incident p-mean of its neighbors, producing a tunable message-passing operator that is averaging for p=2 and nonlinear for other p.
Normalize a learned symplectic recurrent transition using the paper's explicit positive-definite factor instead of projecting the entire matrix onto an orthogonal group. Retain the compact unitary dynamics as the stable transport component and separately damp only the noncompact factor, providing a tunable stability mechanism that preserves symplectic structure.
Replace dense aggregation over patches, points, or graph neighbors with a state-dependent positive quadrature rule. A latent encoder produces the current manifold coordinate z, a small controller chooses a subset of entities, and a convex redistribution step assigns nonnegative weights so that the subset preserves the full aggregation on a learned feature basis. The method is especially suitable for neural operators and vision attention where many spatial entities are redundant along the data…
Construct a sparse directed attention mask over ordered tokens by treating token pairs as edges labeled by time or position. Apply a recursive three-hop spanner construction so that every dense temporal route has a sparse route using at most three times as many hops. Run attention only on the selected edges and use three sparse message-passing layers, or a small learned hop-composition block, to approximate dense causal attention.
Replace repeated high-dimensional quadratic interactions in a neural operator or world model with a reduced latent quadratic map whose basis-pair interactions are precomputed offline. The online computation becomes a small polynomial in latent coefficients, preserving quadratic interaction structure while avoiding repeated full-resolution contractions.
Replace dense attention weights with a compactly supported anisotropic bump derived from the obstacle solution, using one learnable ellipsoid per attention head or feature group. Tokens outside the learned ellipsoid receive exactly zero weight, while tokens inside receive smoothly decaying weights according to a fractional exponent. The learned positive-definite matrix represents orientation, scale, and correlations between feature dimensions.
Replace unconstrained token-to-expert routing with a nonnegative mixture of a small set of feasible routing or communication patterns. Enforce resource limits using the paper's one-sided positive-contribution bound, which gives a conservative certificate without enumerating all joint token activation scenarios. Expand the pattern set only when a separation procedure finds a routing pattern that improves the router objective while adding useful capacity information.
Replace pointwise high-order derivative residuals in an eigenvalue PINN by an assembled dynamic-stiffness residual \(\mathbf W(\omega)q_\theta\), where each element matrix is obtained from homogeneous PDE solutions. The network predicts nodal degrees of freedom or element boundary traces, while the exact frequency-domain operator enforces the physics without differentiating the network multiple times with respect to coordinates.
Insert a sparse heavy-tailed nonlocal mixing operator into a residual sequence, graph, or spatial network so that information can traverse distant positions without stacking many local layers. Use the critical tail exponent s = 1/2, whose truncated first moment grows logarithmically and predicts an effective propagation distance proportional to depth times log depth rather than merely depth.
Replace the dense output of selected linear projections with a two-sided magnitude threshold that emits zero for small values but preserves signed large values. Learn one positive threshold per projection, or optionally one threshold per output channel, so the network discovers where sparse events can be removed while retaining outlier information.
Use cumulative control measures as the input to a neural surrogate instead of the full sequence of control values. For a quasi-static free-boundary system satisfying the paper's average-pressure path-independence assumption, two nonnegative control histories with identical integrals up to time t should produce the same state at t, allowing a smaller training input and fewer distinct control trajectories in the dataset.
Replace a sparse graph layer's separate edge transformations with one joint low-rank factorization of all transformations entering each target node. For target node i, concatenate the neighbor matrices horizontally, project all neighbor features into a shared low-dimensional receiving basis, and reconstruct one output; retain the self transformation exactly. This can reduce edge-parameter storage and message-passing FLOPs when the incoming block row has rapidly decaying singular values.
Replace random node or token subsampling by a subset selected to preserve the averages of low-frequency graph signals exactly or approximately. Build a graph over nodes or tokens, compute its Laplacian eigenspace, and choose a fixed-cardinality subset whose eigenvector moments match those of the full graph; use the retained subset for graph pooling or downstream attention.
Replace recursive adaptive coordinate traversal in a CPPN-based architecture generator with one static multiresolution candidate tensor and an active boolean mask. Evaluate all candidate coordinates in parallel, compute local CPPN-output variance over each parent cell's children, and retain only children whose parent is active and whose predicted variation exceeds a threshold. This preserves adaptive refinement while enabling population batching, JIT compilation, GPU vectorization, and…
Replace online enumeration over a finite action set with a classifier or lookup map whose regions directly return the action minimizing a one-step predictive-control cost. For affine dynamics and quadratic tracking loss, exact action regions are separated by pairwise cost boundaries, so the approximation can be audited against exhaustive predictive control rather than treated as an unconstrained policy.
Model dynamic routing as a multitype branching process: an active token of type d probabilistically creates child activations of type d'. Estimate the corresponding mean offspring operator and regulate its Perron root to a target reproduction rate, typically near one. This should make adaptive-depth or recursively routed networks use sparse computation without producing either rapidly vanishing paths or uncontrolled activation explosions.
Before or during graph-neural-network inference, use a tree dynamic program to choose a limited set of active computation nodes and upgraded message-passing edges. A node receives an embedding from an active landmark only when the selected path has enough residual communication radius, so the planner directly optimizes weighted coverage under a joint node-and-edge budget. The resulting active subgraph is then used by a sparse GNN or graph transformer.
Approximate a dense symmetric interaction matrix in a neural layer by \(\widehat A=C\widehat M C^{\top}\), but compute the small core \(\widehat M\) from a two-sided sketched least-squares fit rather than from the landmark principal submatrix. This preserves signed or indefinite directions and avoids exploding outputs caused by an almost-singular \(A(I,I)\).
Replace repeated substrings in long sequences with nonterminal symbols from an acyclic straight-line grammar, then run the transformer on the compressed sequence. Unlike ordinary fixed tokenization, the compression objective explicitly minimizes the number of reusable binary productions, allowing repeated document-specific or corpus-level motifs to become single units. An expansion map lets the model recover token-level outputs for selected positions.
Construct a sparse fixed orthogonal mixer by repeatedly applying pi/4 rotations to randomly matched pairs of feature coordinates. Place this mixer before top-k feature pruning, sparse projection, or activation quantization so that information is spread across coordinates without using a dense random matrix.
Replace a trainable shallow hidden layer by a deterministic feature dictionary generated from Chebyshev-spaced scalar parameters and quasi-uniform ridge directions. Train only the output linear map, or use the frozen layer as the first stage of a larger network, thereby eliminating hidden-layer backpropagation while retaining a constructive smooth-function approximation guarantee.
Replace dense token-to-token attention in a controlled ablation with a cyclic order-a recurrence whose coefficients are periodic across positions. The resulting mixer has linear cost in sequence length for fixed recurrence order and can enforce a fixed signed periodic boundary condition, providing a compact structured alternative to local attention or a lightweight state-space model.
Replace a naively evaluated mixture of power-law experts with a Newton-envelope layer that computes all monomial magnitudes in log-space and subtracts their maximum before exponentiation. The layer exposes both a stabilized mixture value and soft dominance weights, allowing a downstream MLP to adapt to whichever scaling regime is active without overflow or hand-designed regime splits.
Constrain a graph or token attention layer to behave like a row-stochastic diffusion with walk dimension \(\beta\), and choose its neighborhood radius from the paper's first-hitting-time tail bound. At propagation time \(t\), interactions with effective distance larger than \(r(t,\varepsilon)\) are removed because their total reachability probability is bounded by a prescribed error budget \(\varepsilon\).