✓✓ Beats tuned baseline
2026
Replace dense self-attention by a sparse attention graph whose neighborhoods satisfy the paper's size-dependent expansion condition. This preserves a logarithmically controlled route for every token subset to communicate with the rest of the sequence, reducing quadratic attention cost without allowing disconnected or poorly mixed token groups.
Useful8/10
Difficulty5/10
Novelty5/10
✓✓ Beats tuned baseline
2026
Replace full-state quantized write-back in a deep low-bit residual stack with quantized increment error feedback. The residual branch quantizes the proposed increment after adding the previous carry, while the carry stores the exact discrepancy; this makes the total error telescope instead of accumulating approximately once per layer.
Useful8/10
Difficulty5/10
Novelty6/10
✓✓ Beats tuned baseline
2026
Train a categorical denoiser for the clean token but convert its output analytically into the reverse CTMC jump rates using the exact forward transition kernel. This separates the easy-to-learn clean-token posterior from the quantity required by the reverse process and should keep the uniform-diffusion ELBO finite at initialization, unlike direct denoiser substitution.
Useful8/10
Difficulty4/10
Novelty7/10
✓✓ Beats tuned baseline
2026
Replace dense graph self-attention with two parallel branches: exact softmax attention only over graph neighbors and a global linear-attention branch that summarizes all nodes through feature-space statistics. A learned node-wise gate interpolates between the branches, allowing locally structured nodes to use sparse attention while retaining a global-information path.
Useful7/10
Difficulty4/10
Novelty5/10
✓✓ Beats tuned baseline
2026
Replace ordinary depth-wise feature propagation by a ternary hierarchical block that recursively aggregates three child representations while maintaining separate neutral and defect channels. The block is initialized from the Sierpinski six-vertex recursion, then optionally learns a bounded correction. The neutral channel preserves the paper's cubic mixing law, while the defect channel provides a controlled route for long-range and nonlocal interactions.
Useful7/10
Difficulty6/10
Novelty7/10
✓✓ Beats tuned baseline
2026
Replace dense token-to-token attention by a multiscale spiderweb communication pattern. Tokens first aggregate upward through a dyadic hierarchy, communicate horizontally only with a small number of cells at the appropriate height, and then receive information broadcast downward. Hyperbolic distance supplies a principled rule for choosing the height at which two tokens interact: nearby tokens interact at fine scales, while far-apart tokens interact through coarse representatives.
Useful7/10
Difficulty5/10
Novelty6/10
✓✓ Beats tuned baseline
2026
Represent each input graph by a short ordered sequence of topological tokens rather than a persistence diagram. For each overlapping interval on a shared filtration grid, compute the number of connected components and independent one-dimensional cycles, concatenate these counts with inexpensive graph summary statistics, and process the resulting sequence with a small Transformer. The representation should retain multiscale structural information while avoiding pairwise persistence-diagram…
Useful7/10
Difficulty6/10
Novelty6/10
✓✓ Beats tuned baseline
2026
Replace unconstrained softmax cross-attention with a many-to-many transport matrix whose row and column masses have explicit upper bounds. Compute the attention cost from both feature similarity and pairwise relational disagreement, so a token is attended to only when its relationships to other tokens are jointly compatible. The inequality constraints provide a principled dustbin-free mechanism for ignoring distractor tokens.
Useful7/10
Difficulty6/10
Novelty5/10
✓✓ Beats tuned baseline
2026
Construct a residual sequence or depth network whose nonnegative influence density follows a discretized noisy Fisher-KPP equation: local influence diffuses, grows when small, saturates at a finite carrying capacity, and receives state-dependent noise. Use this density to gate ordinary feature updates rather than relying only on unconstrained residual additions. The mechanism predicts a measurable propagation speed and an instability boundary, allowing the architecture to be falsified…
Useful7/10
Difficulty6/10
Novelty8/10
✓✓ Beats tuned baseline
2026
Replace or augment a recurrent or state-space block with a small ensemble of stable linear memory lifts. Each lift stores a low-dimensional state whose repeated matrix powers generate a structured long-range convolution, and a learned gate mixes the experts using both their current predictions and their slowest block-operator mode. The module can represent multiple memory timescales without explicitly storing a long token history.
Useful7/10
Difficulty6/10
Novelty5/10
✓✓ Beats tuned baseline
2026
Replace dense pairwise interactions between all forecast horizons with nested time-shell summaries. For sorted horizons, the readout at shell j receives a cumulative embedding of all coefficients or queries assigned to later horizons, reproducing the paper's dependence on products such as \(\Pi_j=\prod_{l>j}e^{\alpha_l}=e^{\sum_{l>j}\alpha_l}\). This gives an \(O(Kd)\) multi-horizon interaction instead of an \(O(K^2d)\) temporal attention block and should work best for weak-memory…
Useful7/10
Difficulty5/10
Novelty8/10
✓✓ Beats tuned baseline
2026
Route tokens to experts using Laguerre cells defined by the minimum control energy needed to move a token embedding to each expert prototype, rather than by Euclidean distance or an unconstrained learned router logit. Per-expert dual weights deform the cells so that minibatch routing follows prescribed expert capacities, giving a geometrically interpretable alternative to auxiliary load-balancing losses.
Useful7/10
Difficulty6/10
Novelty6/10
✓✓ Beats tuned baseline
2026
Replace a recurrent token-mixing operation with a finite-state carrier scan over binary or quantized token features. The local transition table is constrained to conserve a scalar token weight and to be bijective, making the mixer reversible, constant-memory, and less prone to activation drift than a generic recurrent layer.
Useful6/10
Difficulty5/10
Novelty8/10
✓✓ Beats tuned baseline
2026
Build an input-conditioned unitary transformation as an ordered product of exponentials of anti-Hermitian matrices, with each factor controlled by a univariate function of one input coordinate or one learned scalar projection. This replaces a dense multivariate matrix-valued controller with separable scalar nonlinearities while preserving exact unitarity at every forward pass.
Useful6/10
Difficulty6/10
Novelty7/10
✓✓ Beats tuned baseline
2026
Replace random Fourier features or a dense sinusoidal positional encoding with a compact bank whose frequencies are the continued-fraction denominators of an irrational number. Inverse-frequency amplitudes provide multiscale structure with a controlled sub-Lipschitz regularity profile, while lacunarity reduces the number of frequencies needed to represent oscillatory structure.
Useful6/10
Difficulty3/10
Novelty5/10
✓✓ Beats tuned baseline
2026
Estimate the temporal spectrum of each sequence channel using a locally private procedure, then apply a regularized inverse-square-root spectral filter before the sequence enters attention or an SSM. The filter removes predictable low-frequency or narrow-band redundancy while avoiding unstable amplification at frequencies where the private estimate is small.
Useful6/10
Difficulty5/10
Novelty7/10