✗ Failed on benchmark
2026
Construct a graph of cheap prediction agents or reasoning traces and use a sparse set of expensive verifier calls as graph anchors. Select the next verifier location by the exact reduction in a trace-inverse coherence objective per unit cost, rather than by uncertainty or random sampling. This creates a budgeted mixture-of-agents architecture that can spend computation where it most improves global consensus.
Useful7/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Augment an optimizer with a measurable redistribution time for its internal state and compare it with the time scale of the changing gradient field. Use the resulting Damkohler number to interpolate between a fast quasi-static preconditioner and a history-preserving, non-equilibrium update, rather than applying one optimizer regime throughout training.
Useful7/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Replace isotropic parameter penalties and diagonal Fisher estimates with a task-covariance interference budget. The update is damped only in directions where old-task features have large variance, while directions absent from old-task feature support remain available for learning the new task. This may preserve old-task performance with less loss of plasticity than unconditional projection.
Useful7/10
Difficulty5/10
Novelty6/10
✗ Failed on benchmark
2026
Train a neural state-space model using all replayed transitions, but assign larger weights to samples near the current operating context rather than discarding distant samples. Add a strictly positive weight floor so local adaptation cannot eliminate global coverage or make the regression problem rank-deficient. This should improve prediction across nonlinear regimes while retaining the numerical robustness of full-data training.
Useful7/10
Difficulty4/10
Novelty5/10
✗ Failed on benchmark
2026
Use the reachability verifier as an optimization controller: permit a neural controller update only when the proposed parameter step remains inside a certified STL-safe trust region, and shrink the region when the reachable robustness margin collapses. This turns verification from an expensive final check into feedback that prevents gradient descent from crossing a temporal-logic feasibility boundary.
Useful7/10
Difficulty7/10
Novelty8/10
✗ Failed on benchmark
2026
Replace an unconstrained second-order residual or state-space block with a position-velocity system whose damping is the gradient or subgradient of a convex function. Compute the next state implicitly, so the damping cannot inject energy and the resulting layer is robust to large learned damping nonlinearities, nonsmooth activations, and long rollouts.
Useful7/10
Difficulty5/10
Novelty5/10
✗ Failed on benchmark
2026
Treat optimizer configurations as elements of a finite intervention poset and decompose validation loss or training traces into pure causal effects rather than raw ablation differences. The recovered second- and higher-order effects reveal whether, for example, momentum and adaptive preconditioning are complementary, redundant, or destabilizing, and can be used to select a smaller optimizer or construct a better configuration.
Useful7/10
Difficulty4/10
Novelty7/10
✗ Failed on benchmark
2026
Attach a sampling-based rollout correction head to a neural policy or learned world model, and adapt its temperature and number of rollouts so that approximation error stays within the contraction margin of a nominal policy. The controller should spend samples only when the local state-dependent error gain is close to violating the small-gain condition, instead of using a fixed MPPI sample count everywhere.
Useful7/10
Difficulty6/10
Novelty7/10
✗ Failed on benchmark
2026
Replace an opaque adaptive-optimizer state update with a small controller variable obtained by minimizing a strongly convex energy jointly associated with the proposed parameter motion. The controller is allowed to relax toward the current gradient before the parameter update, while the visible update uses the reduced energy and its envelope gradient. This creates an optimizer whose hidden geometry is optimized rather than inherited from a fixed exponential-moving-average recurrence.
Useful7/10
Difficulty6/10
Novelty6/10
✗ Failed on benchmark
2026
Replace the fixed numerical stabilizer in signSGD by an exponentially decaying stability path, so the optimizer remains sign-like for a controllable duration instead of eventually reverting toward ordinary gradient descent as gradients become small. Sweep the decay rate as an explicit implicit-bias parameter: slower annealing should retain the non-Euclidean, barrier-like bias, while faster annealing should approach the sign endpoint more closely.
Useful7/10
Difficulty4/10
Novelty6/10
✗ Failed on benchmark
2026
Treat every low-rank basis refresh as a change of coordinates instead of assuming that old optimizer coordinates remain aligned with the new basis. Transport the first moment with the basis-overlap matrix, but collapse the second moment to a rotation-blind isotropic estimate rather than applying the same coordinate transformation to elementwise squared moments. This should eliminate second-moment staleness while preserving the memory savings of low-rank optimization.
Useful7/10
Difficulty4/10
Novelty7/10
✗ Failed on benchmark
2026
Use a smoothed Burg entropy as the mirror map in a proximal-gradient optimizer for positive or simplex-valued neural parameters. The optimizer performs a Bregman-proximal step instead of an additive Euclidean update, while the smoothing parameter avoids the singularity of ordinary Burg entropy at zero.
Useful7/10
Difficulty5/10
Novelty5/10
✗ Failed on benchmark
2026
Replace the sign-flip-only dynamics of high-index saddle search with low-rank inverse-curvature scaling on the estimated negative-curvature subspace. Directions with small negative Hessian eigenvalues then receive approximately curvature-independent updates instead of extremely slow updates proportional to their tiny curvature.
Useful7/10
Difficulty6/10
Novelty7/10
✗ Failed on benchmark
2026
Add a latent mode bank whose coordinates are learned by neural power iteration on observed state transitions rather than by jointly fitting an unconstrained latent dynamics model. Each mode is repeatedly regressed toward its one-step pushforward, normalized under the data distribution, and deflated against previously learned modes. The resulting latent coordinates are constrained to have approximately linear, diagonal dynamics, which should improve long-horizon prediction and make the…
Useful7/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Use the M phase-aligned parameterizations produced by cyclic reformulation as an empirical ensemble of neural dynamics rather than selecting one phase or averaging only predictions. Their centroid supplies a nominal model, while their convex hull defines a low-dimensional uncertainty set used for robust rollout training and uncertainty-aware inference.
Useful6/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Train a neural dynamics predictor or policy output head against an empirical buffer of observed prediction-error scenarios rather than only nominal targets. For each input, require the predicted output plus every sampled residual trajectory to remain inside the admissible set, using an exact nonnegative slack penalty when robust feasibility is impossible. This should reduce rare but operationally important constraint violations while preserving nominal tracking accuracy.
Useful6/10
Difficulty5/10
Novelty6/10
✗ Failed on benchmark
2026
Attach a model-free residual-dynamics observer to a neural multi-step forecaster. Instead of asking the network to relearn persistent periodic or autoregressive disturbances, maintain a Hankel dictionary of recent forecast errors and use ridge reconstruction to predict the next residual sequence online. Add the predicted residual to the network forecast with a confidence-dependent correction gain.
Useful6/10
Difficulty4/10
Novelty7/10
✗ Failed on benchmark
2026
Track an exponentially discounted approximation to the current min-max saddle gap and use it to control the optimizer of a GAN or adversarial learner. If the recent gap rises, reduce both players' step sizes and clear stale momentum; if it falls consistently, cautiously increase the step sizes. Unlike ordinary loss EMAs, this signal measures whether each player is close to a recent best response and can detect equilibrium-tracking failure even when generator and discriminator losses look benign.
Useful6/10
Difficulty4/10
Novelty7/10
✗ Failed on benchmark
2026
Use the paper's negative-semidefinite interaction curvature to detect and compensate for destructive coupling among layerwise learning-rate, momentum, or preconditioner mechanisms. Instead of independently tuning mechanism amplitudes, estimate their reduced curvature after hidden optimizer states relax, then apply a low-rank trust-region step or freeze mechanisms whose interaction curvature is too negative.
Useful6/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Replace a fixed soft-threshold, ReLU-like gate, or manually chosen activation shrinkage with a monotone learned shrinkage function fitted by an observed-data quadratic-risk criterion. The gate can interpolate between identity, ridge-like attenuation, hard thresholding, and lasso-like soft thresholding, allowing each layer or channel group to adapt its bias–variance tradeoff from the current minibatch.
Useful6/10
Difficulty5/10
Novelty6/10
✗ Failed on benchmark
2026
Insert an active-set reduction step into a binary energy layer or Hopfield-style discrete optimizer. Coordinates whose signs are stable and whose local fields have a rigorous margin are frozen, while their interactions are folded into an induced bias and only the unresolved tail is updated. This preserves the exact conditional quadratic objective and can reduce dense interaction cost substantially when the state becomes polarized.
Useful6/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Insert a scalar flux-correction-style limiter after a neural operator predicts a conservative state or residual. Interpolate between a known-admissible baseline state and the learned high-order candidate, choosing the largest coefficient that satisfies a geometric family of linear inequalities encoding positive density, positive pressure, and subluminal velocity. This retains as much of the neural prediction as possible instead of independently clipping physical variables.
Useful6/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Estimate local curvature, third derivative, and gradient-noise variance, then compensate for the stationary displacement predicted by the paper rather than assuming client averaging removes all bias. The first implementation should operate coordinatewise on a one-dimensional or diagonal quadratic-plus-cubic federated objective, where the paper's coefficient has a direct interpretation.
Useful6/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Replace Gaussian perturbations in a low-dimensional neural-network optimizer with independent double-geometric integer mutations and adapt each mutation scale using its exponential-family natural gradient. Apply the method to layerwise quantization scales, adapter coefficients, pruning thresholds, or other integer/discrete hyperparameters rather than to every individual weight.
Useful6/10
Difficulty5/10
Novelty7/10