✗ Failed on benchmark
2026
Replace the global EMA update for each linear-layer momentum matrix with a delta-rule update that learns the current output-side gradient value only along the current input-key direction. Frequently occurring directions are corrected repeatedly, while rarely visited directions are not unnecessarily overwritten or uniformly decayed. Use the resulting matrix as the ordinary momentum buffer in SGD, AdamW, or another optimizer.
Useful8/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Replace an instantaneous diagonal optimizer with a causal convolution of recent gradients, where cross-layer or cross-module gradient correlations define a finite-memory Onsager response matrix. Estimate the response at several parameter-block pairs and lags, integrate it to obtain a finite-time transport matrix, and use its regularized inverse or symmetric part to precondition the update. This targets optimization regimes in which gradients propagate between blocks with measurable delay, such…
Useful8/10
Difficulty6/10
Novelty7/10
✗ Failed on benchmark
2026
Attach a cheap risk score to each neural-network prediction and skip an expensive verifier, ensemble, diffusion refinement, retrieval call, or human review when the score is below a calibrated threshold. Independently audit a random subset of skipped examples using the expensive ground-truth procedure, and select the largest skip threshold whose exact confidence bound keeps the violation rate below a target budget.
Useful8/10
Difficulty4/10
Novelty7/10
✗ Failed on benchmark
2026
Compress each hidden layer by retaining directions that are simultaneously reachable from the observed input distribution and observable at the network output. Unlike PCA or SVD, the retained subspace is weighted by downstream task sensitivity, so high-variance but output-irrelevant directions can be removed while low-variance predictive directions are preserved.
Useful8/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Insert a linear Johnson–Lindenstrauss bottleneck around a set of jointly processed representations, choosing its width from the sharp finite-set dimension bound rather than from the model's nominal hidden size. The projection should preserve pairwise distances between tokens, patches, or retrieved items, allowing a downstream attention or MLP block to operate at lower width while retaining the geometry relevant to similarity computations.
Useful7/10
Difficulty5/10
Novelty5/10
✗ Failed on benchmark
2026
Replace the usual momentum schedule in a neural-network optimizer with a discretization of the paper's lemniscate-acceleration ODE. The method uses a time-dependent friction coefficient that is initially very large and then decays according to lemniscate sine and cosine functions, targeting faster reduction of the gradient norm than constant-momentum SGD or standard Nesterov schedules.
Useful7/10
Difficulty5/10
Novelty8/10
✗ Failed on benchmark
2026
Replace a static batch size with a schedule optimized for a prescribed learning-rate schedule and a fixed total number of processed examples. Steps whose stochastic-gradient noise has a large effect on the paper's loss bound receive larger batches, with the weighting determined by the future learning-rate tail rather than by a hand-designed warmup or cooldown rule.
Useful7/10
Difficulty5/10
Novelty5/10
✗ Failed on benchmark
2026
Replace a neural-network optimizer's globally fixed learning-rate geometry with an adaptive quadratic trust region. At every update, construct a local curvature model, accept or reject the step using the ratio between realized and predicted loss decrease, and expand or contract the radius accordingly; the same controller should automatically become conservative in nonconvex regions and Newton-like near a well-conditioned minimum.
Useful7/10
Difficulty6/10
Novelty6/10
✗ Failed on benchmark
2026
Construct overlapping attention windows from a block design instead of using one dense sequence-by-sequence attention matrix. Every token appears in exactly $r$ windows and every token pair co-occurs in exactly $\lambda$ windows, giving uniform coverage and avoiding the uneven connectivity of arbitrary sparse masks.
Useful7/10
Difficulty5/10
Novelty6/10
✗ Failed on benchmark
2026
Replace a capacity-penalty-only MoE router with a nonnegative shadow price for each expert, capacity bucket, or hardware resource. Route each token using predicted utility minus the relevant price, while computing a decomposed optimistic objective that certifies how much utility remains above the feasible routed value.
Useful7/10
Difficulty4/10
Novelty5/10
✗ Failed on benchmark
2026
Compress a transformer KV cache by selecting actual past tokens whose key or hidden-state columns form a stable basis for all cached tokens. Instead of retaining tokens with the largest attention scores or leverage scores independently, compute rank-revealing pivoting of the leading right-singular-vector matrix and retain its pivot columns, then evaluate attention using the representatives plus an optional low-cost residual correction.
Useful7/10
Difficulty5/10
Novelty6/10
✗ Failed on benchmark
2026
Partition tokens into a rectangular grid and use Latin-square labels to define several sparse attention heads. Each head attends only to tokens sharing a row, column, or Latin label, while orthogonality guarantees that every pair of labels occurs at most once, reducing systematic blind spots and repeated collisions. The resulting masks are deterministic, reusable across examples, and can be generated without learned routing scores.
Useful7/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Regularize the local recurrent Jacobian by its spectral radius rather than imposing the overly conservative operator-norm condition $\|J\|_2<1$. This permits useful non-normal updates with transient amplification while explicitly pushing the asymptotic dynamics toward a stable fixed point.
Useful7/10
Difficulty7/10
Novelty7/10
✗ Failed on benchmark
2026
Replace a weight-tied residual or neural-ODE stepper with an explicit Runge–Kutta method satisfying the reused-last-stage conditions. The final derivative is evaluated at the exact endpoint and becomes the first derivative of the next step, saving one expensive neural-vector-field call per step while preserving the designed integration order.
Useful7/10
Difficulty5/10
Novelty5/10
✗ Failed on benchmark
2026
Construct a sparse token-interaction graph from attention affinities and recursively split it by removing the closed neighborhoods of a small set of separator tokens. Separator tokens retain global communication, while the resulting anti-adjacent components perform local attention independently, reducing quadratic attention and KV-cache costs. The induced Erdos-Pósa theorem supplies a structural diagnostic: graphs with few anti-adjacent long-cycle or theta packings should admit small…
Useful6/10
Difficulty7/10
Novelty8/10