Unverified
2026
Construct sparse attention masks as regular or almost-regular hypergraphs: every query selects exactly w keys, while every key receives nearly equal incidence within each score or locality class. This preserves a fixed sparsity budget while removing systematic positional and head-level load imbalance that can cause hot keys, uneven memory traffic, and unstable gradients.
Useful6/10
Difficulty6/10
Novelty6/10
Unverified
2026
Condition a temporal neural network on a tempo or dilation ratio through a homomorphism from multiplicative positive scales to additive latent shifts. A ratio composed from several scale changes then produces the sum of their learned effects, allowing interpolation and extrapolation to rates absent from training instead of using an independent embedding per rate.
Useful6/10
Difficulty4/10
Novelty5/10
Unverified
2026
Augment a causal sequence model with a small hierarchy of prefix summaries weighted by powers of the logarithmic rank of each preceding token. The summaries retain order-sensitive deviations from a baseline representation while costing O(KNd) for sequence length N, hierarchy width K, and hidden dimension d, instead of O(N^2d) dense attention.
Useful6/10
Difficulty4/10
Novelty6/10
Unverified
2026
Replace one-shot softmax expert routing with a small number of synchronous routing rounds in which each token resamples an expert with probability proportional to that expert's current load raised to a power \(\alpha>1\). The resulting positive feedback rapidly creates a dominant routing basin, potentially reducing the number of active experts and communication groups at inference while retaining a controllable exploration phase through the initial round or a token-specific score factor.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Augment ordinary edge or line-integral features with a learned second-level area term, then train the model to make predictions invariant to path subdivision and stable under nearby path perturbations. The second-level feature captures ordered pair interactions that a sum of edge increments cannot represent, providing a compact alternative to explicitly processing every intermediate point in a trajectory or graph walk.
Useful6/10
Difficulty4/10
Novelty5/10
Unverified
2026
Give each query-token pair a positive adaptive edge weight that evolves by a multiplicative rule instead of relying only on instantaneous dot-product attention logits. Edges whose aggregate interaction is useful can grow, while overloaded or incompatible neighborhoods can shrink. Sparse initialization is preserved because an edge initialized at zero remains zero under the multiplicative dynamics.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Replace independent binary early-exit or token-pruning decisions with a monotone randomized survival process for each token or expert route. A token can lose survival mass at each layer but cannot become active again; the model is trained with a reflected obstacle-style penalty that activates when the predicted value of continuing computation is below the value of stopping plus the compute cost. Mean-field statistics are computed over currently surviving tokens, making routing less sensitive to…
Useful6/10
Difficulty5/10
Novelty5/10
Unverified
2026
Add a learned scalar ordering to a directed graph attention layer and retain only forward edges, producing a DAG attention mask without requiring a supplied topological order. Train the ordering with a differentiable surrogate for weighted surplus, and regularize it toward the paper's explicit half-weight-minus-l2 certificate. This supplies a principled alternative to random masking or unconstrained bidirectional graph attention when causal or hierarchical information flow is desirable.
Useful6/10
Difficulty5/10
Novelty8/10
Unverified
2026
Replace a conventional softmax router or fixed halting score with a scalar confidence state that evolves as a bounded martingale diffusion. The state starts at the network's prior confidence, receives evidence-dependent stochastic increments, and is absorbed at 0 or 1; absorption selects an MoE expert or halts additional transformer blocks. State-dependent volatility lets the model explore aggressively when uncertain and commit rapidly when confident, while the martingale constraint prevents…
Useful6/10
Difficulty5/10
Novelty7/10
Unverified
2026
Replace fixed sequence-to-sequence attention with a dynamically maintained tree of connected token groups. Groups can be fused to reduce the number of attention units or split when their representation is heterogeneous, while hypergraph connectivity and nestedness ensure that every intermediate hierarchy remains valid.
Useful6/10
Difficulty6/10
Novelty5/10
Unverified
2026
Replace unconstrained MoE router logits with structured phase scores indexed by N-subsets of M ordered parameters. Each token is assigned to the dominant phase, while neighboring routing regions obey the Grassmannian rule that adjacent labels share N-1 indices, reducing arbitrary fragmented decision boundaries and encouraging smooth expert transitions.
Useful6/10
Difficulty5/10
Novelty7/10
Unverified
2026
Replace unconstrained pairwise token grouping with a tree whose edges carry independent merge or cut variables. The connected components of the retained edges define a valid partition at every forward pass, while learned edge gates control the amount of token aggregation. A coarse component-level computation can then replace part of dense attention.
Useful6/10
Difficulty5/10
Novelty7/10
Unverified
2026
Compute separate doubled-angle orientation order parameters for left and right image regions, then expose their sum and difference as symmetric and antisymmetric global features. This gives a network a low-dimensional inductive bias for global vertical structure versus left-right imbalance, while retaining magnitude channels that indicate when either readout is undefined because orientations cancel.
Useful6/10
Difficulty5/10
Novelty7/10
Unverified
2026
Replace a dense block of a large neural-network weight matrix with an adaptively constructed cross approximation, but prevent premature termination using residual checks on O(m+n) additional diagonal-like entries. Accept a rank only after the probe residual has remained below tolerance for several consecutive iterations; otherwise continue adding pivots.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Equip an autoregressive model with a hard constraint that generated token ranks cannot contain a nondecreasing subsequence of length k. Maintain a patience-sorting-style state representing the smallest ending token rank achievable by subsequences of lengths 1 through k-1, mask tokens that would create a length-k subsequence, and optionally sample uniformly among valid continuations using exact suffix counts. This provides a mathematically controlled data-augmentation or decoding regime for…
Useful6/10
Difficulty5/10
Novelty7/10
Unverified
2026
Replace or augment a residual neural layer with a Fourier-domain scale-selective flow containing a learned second-order term and a fourth-order stabilizer. The block permits controlled low-frequency amplification, as required by the KS infrared mechanism, while damping high-frequency feature noise and preventing unbounded spectral growth.
Useful6/10
Difficulty5/10
Novelty7/10
Unverified
2026
Require the calibrated state posterior to remain unchanged when evidence is presented through information-equivalent prompt templates. Compare state distributions after semantic aggregation rather than raw token probabilities, and add a total-variation consistency penalty during calibration or fine-tuning.
Useful6/10
Difficulty4/10
Novelty6/10
Unverified
2026
Add a continuation-value correction and an explicit positive margin to routing or adaptive-computation decisions that consume a stateful resource. The margin prevents the degeneracy in which learned rent exactly cancels immediate surplus, making the network highly sensitive to noise, quantization, or small state-estimation errors.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Construct attention groups as hyperedges of a linear r-uniform hypergraph: every pair of tokens is allowed to share at most one group, while each token participates in approximately the same number of groups. Apply local attention inside each group and aggregate the outputs across groups. The construction inherits the paper's sharp capacity bound and prevents both redundant pair interactions and high-degree token hubs.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Replace a learned embedding for each integer with a compositional embedding of its variable-length Gray codeword. Add an auxiliary constraint that numerically adjacent values have nearby representations, while preserving the ordinary task loss so that the model can learn when numerical adjacency matters.
Useful6/10
Difficulty4/10
Novelty8/10
Unverified
2026
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
Unverified
2026
Replace an unconstrained relative-position attention-bias table by a learnable two-dimensional Fourier representation, or regularize the bias toward a small Fourier ratio. The ratio favors coefficient concentration without depending on the absolute scale of the bias, allowing the trained bias to be reconstructed from a small number of dominant frequencies.
Useful6/10
Difficulty4/10
Novelty7/10
Unverified
2026
Wrap a nominal gradient-based optimizer with a diagonal sign matrix that flips updates independently for parameter blocks, while a scheduler tests candidate sign configurations using short-horizon decrease of a Lyapunov-like training energy. The wrapper never changes the magnitude of the nominal update, and when the effective sign pattern is constant, it should recover the behavior of the correctly oriented nominal optimizer after a finite search period.
Useful6/10
Difficulty5/10
Novelty8/10
Unverified
2026
Assign separate sharpness or temperature parameters to two nonlinear subnetworks and anneal them according to a directional chart instead of driving both to their singular limits at the same rate. The optimizer explicitly tracks the ratio of the two scales and changes the schedule when the local Jacobian approaches a stability or bifurcation boundary. This tests whether the order and relative rate of sharpening, rather than only the final activation shape, controls optimization stability and…
Useful6/10
Difficulty5/10
Novelty8/10