△ Mechanism confirmed, baseline not beaten
2026
Turn an iterative optimization or equilibrium computation inside a neural network into a differentiable layer whose backward pass solves the implicit adjoint system with conjugate gradients or GMRES using only automatic-differentiation matrix-vector products. This avoids storing unrolled iterations and avoids explicit Hessian or Jacobian construction, enabling longer solver horizons and lower-memory implicit architectures.
Useful8/10
Difficulty6/10
Novelty5/10
△ Mechanism confirmed, baseline not beaten
2026
Before quantizing a matrix product, reparameterize its factors as A'=AT and B'=T^{-1}B, preserving the exact full-precision product while changing the quantization difficulty of each factor. Choose a positive diagonal T=diag(t_1,...,t_K) that minimizes predicted post-quantization product error, rather than using output-channel scaling or a fixed heuristic grid. The gauge can be shared across several products when transformed-copy cost matters.
Useful8/10
Difficulty5/10
Novelty6/10
△ Mechanism confirmed, baseline not beaten
2026
Replace Euclidean or entrywise Kronecker fitting of a layer curvature matrix with its affine-invariant projection onto G = A tensor B. Use the resulting factors as a compact SPD preconditioner in the optimizer, while solving the projection through logarithmic residual partial traces and Armijo line search.
Useful8/10
Difficulty6/10
Novelty6/10
✗ Failed on benchmark
2026
Compress a trained wide analytic-activation MLP by fitting a narrow same-depth student to the teacher's function values and input derivatives, rather than matching only outputs on a calibration dataset. Choose the student width from the input dimension and target error, with a target scaling m = O((log(1/epsilon))^d_in), and use sequential layer fitting plus channel reweighting to limit error accumulation through depth.
Useful8/10
Difficulty6/10
Novelty7/10
△ Mechanism confirmed, baseline not beaten
2026
Train a fixed-rank neural weight update Y=USV^T with a projector-splitting Runge–Kutta step instead of independently applying Adam or gradient descent to U, S, and V. The update evolves the full low-rank matrix using the neural gradient but performs QR-based factor updates, avoiding S^{-1} and remaining stable when adapter singular values collapse or cross zero. Use a common-base midpoint construction so every internal stage starts from the same U,V basis and remains rank r.
Useful7/10
Difficulty5/10
Novelty6/10
✓✓ Beats tuned baseline
2026
Replace a stack of independently parameterized residual or MLP blocks with a small latent grid or vector repeatedly updated by one shared transition rule. Let the number of updates depend on the current latent state, so easy examples terminate early while hard examples receive more computation, potentially improving parameter efficiency and extrapolation.
Useful7/10
Difficulty5/10
Novelty6/10
△ Mechanism confirmed, baseline not beaten
2026
Parameterize a trainable weight update as \(\Delta W=UV^{\top}\) with an excessive initial rank \(r\), and penalize active columns using an exact column \(\ell_{2,0}\) penalty. Increase \(\lambda\) along a warm-started path and hard-delete redundant paired columns, producing an automatically selected rank without training a separate model for every candidate rank. Apply scale balancing after each update so pruning decisions are invariant to reciprocal rescaling of factor pairs.
Useful7/10
Difficulty5/10
Novelty6/10
Audited (legacy)
2026
Replace uniform embedding dimensions with a globally budgeted allocation based on the estimated spectral complexity of each categorical feature. Tables whose category representations have large leading singular-value energy receive more dimensions, while high-cardinality tables are penalized because each extra dimension consumes more parameters.
Useful7/10
Difficulty4/10
Novelty6/10
✗ Mechanism failed
2026
Replace an opaque MLP vector field with a stack of trainable symbolic primitives that can express linear terms, monomials, products, and related analytic operations. Apply an L1 penalty and prune small primitive coefficients after rollout training, yielding a compact dynamics module that is cheaper to evaluate and easier to inspect.
Useful7/10
Difficulty6/10
Novelty6/10
Unverified
2026
Add a neural feature, adapter, or expert block only through the component of its outputs that is orthogonal to the span of all previously installed blocks. Quotient coefficient directions that produce nearly identical outputs with an SVD or pseudoinverse, so the new block contributes intrinsic representational dimensions instead of duplicating old features. The expected benefit is a smaller effective architecture and better-conditioned block expansion.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Replace globally backpropagated hidden-layer losses with independent recurrent layers that receive bottom-up, top-down, and lateral inputs. Train each layer to assign low activity to correctly paired input-label examples and high activity to mismatched examples, then classify by selecting the label with the lowest accumulated surprise.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Use enumerated weighing matrices as sparse orthogonal channel-mixing operators inside MLPs or residual blocks. Their ternary entries reduce multiplication to signed additions, while exact orthogonality prevents amplification or attenuation of feature norms; a trainable fixed-support version can recover expressivity without giving up computational sparsity.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Represent a neural-network weight tensor by rank-one terms whose mode factors are selected from shared orthonormal bases, and impose the same basis alignment across tensor flattenings. During or after training, retain the largest coefficients to obtain a structured truncation analogous to truncated SVD. This should produce better-conditioned tensorized layers than unconstrained CP factors while preserving a directly controllable accuracy/compute tradeoff.
Useful6/10
Difficulty5/10
Novelty5/10
Unverified
2026
Replace heuristic magnitude pruning in a layerwise convex reconstruction problem with safe screening based on a perspective relaxation of the cardinality constraint. A channel is removed only when a lower bound for every solution containing that channel exceeds the loss of a feasible incumbent; conversely, a channel is forced to remain when every solution excluding it is provably worse.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Train a square orthogonal neural mixer while maximizing its entrywise fourth-power concentration. When optimization reaches a non-permutation stationary configuration, explicitly test rank-two row or column rotations and take a rotation with positive exact second variation, using the paper's constructive saddle-escape mechanism.
Useful6/10
Difficulty5/10
Novelty6/10
Unverified
2026
Replace a wide activation vector or spatial feature field by a small set of weighted coordinates that preserves the p-norm of every activation in a learned low-dimensional subspace. Unlike ordinary pruning, the selection objective is uniform over the whole coefficient sphere, so the compressed representation is designed to preserve unseen linear combinations and not merely the training examples.
Useful6/10
Difficulty6/10
Novelty7/10
✓✓ Beats tuned baseline
2026
Add an online low-rank reconstruction layer before a recommender or ranking MLP. It estimates a latent action-feature subspace from masked candidate vectors, freezes that subspace for an epoch, reconstructs each partially observed candidate in the latent coordinates, and feeds only those coordinates to the predictor. The method is most promising when the ambient candidate dimension d is large but the effective rank m is small and missingness is not too severe.
Useful6/10
Difficulty5/10
Novelty7/10
✓✓ Beats tuned baseline
2026
Use an approximate decision diagram to select a structured subset of neurons, channels, attention heads, or attention edges when their quadratic interactions are sparse or inverse-sparse. Merge states that agree on a local interaction boundary and accept a tunable epsilon loss in the pruning objective, obtaining a representation whose size is linear in model width for fixed accuracy tolerance.
Useful6/10
Difficulty6/10
Novelty7/10
Unverified
2026
Replace ordinary per-coordinate activation normalization or covariance whitening with a learned or frozen affine map based on a minimum-volume ellipsoid calibrated to the activation convex hull. Estimate the hull's Minkowski symmetry and use the bound \(\rho=\sqrt{d/\alpha}\) to set a principled clipping radius, where \(d\) is feature dimension and \(\alpha\) is the measured symmetry. This should reduce worst-case clipping of asymmetric activation distributions without requiring a much larger…
Useful5/10
Difficulty6/10
Novelty4/10
Unverified
2026
Replace a collection of dense task-specific linear layers with a common sparse structural matrix and task-specific edge strengths. All tasks share the same learned connectivity pattern, but retain independent values on active connections, allowing parameter sharing without forcing identical interactions.
Useful5/10
Difficulty5/10
Novelty5/10
Unverified
2026
Build an implicit layer from a piecewise-linear maximal monotone operator on visible variables z_* and auxiliary variables z_**, then eliminate the auxiliary block rather than exposing it in the network output. Compute the layer through a fixed point of the eliminated component of a nonexpansive resolvent, with damping when the auxiliary map is not strictly contractive.
Useful5/10
Difficulty6/10
Novelty6/10
Unverified
2026
Replace a neural layer's dense parameters by a real parameter tensor and a binary activity mask, then update the mask using importance scores measured across multiple perturbation intervals. Gates that repeatedly occur in non-dominated accuracy–cost configurations receive larger scores and are more likely to be retained or activated, avoiding uniformly random sparse rewiring.
Useful5/10
Difficulty5/10
Novelty4/10