✗ Failed on benchmark
2026
Replace a fixed optimizer memory order with a nested family of gradient-integral controllers. Training begins with a first-order update and activates additional accumulated-gradient states only after an exponentially smoothed residual fails to decrease for several decision intervals; newly activated gains are ramped from zero, so the parameter update remains continuous and previously learned states are preserved. The optimizer should use little memory on easy problems and acquire longer memory…
Useful8/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Replace dense attention over structured object tokens with attention over role-filler tensor-product representations. A learned query specifies both a role and a filler, retrieves objects matching that binding, extracts a target role, and rebinds the extracted filler into an output object.
Useful8/10
Difficulty5/10
Novelty6/10
✗ Failed on benchmark
2026
Replace derivative-based latent-dynamics fitting with an integral regression and maintain a history stack selected by the smallest eigenvalue of its information matrix. The model should perform aggressive parameter updates only when the estimated latent regressors are sufficiently exciting, while a perturbation bound prevents false excitation caused by inaccurate hidden-state estimates.
Useful8/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Add a controllable delay to the gradient force during optimization so that parameters follow a delayed-gradient dynamical system. Choose the delay below the stability boundary for ordinary training, and temporarily cross the boundary when the optimizer is trapped in a sharp or stagnant basin, causing stochastic fluctuations to be amplified out of the basin rather than waiting for a rare Arrhenius escape.
Useful8/10
Difficulty6/10
Novelty7/10
✗ Failed on benchmark
2026
Replace single-Gaussian uncertainty propagation in a neural state-space or world model with a finite mixture of Gaussian latent states. Each component is propagated through the learned nonlinear dynamics, and components are merged or pruned only when their Wasserstein discrepancy is below a prescribed tolerance, preserving multimodal futures while keeping computation bounded.
Useful8/10
Difficulty6/10
Novelty7/10
✗ Failed on benchmark
2026
Augment an RNN or state-space model with a three-dimensional auxiliary spin updated by noncommuting rotations associated with event types or token classes. The ordered product preserves information that additive counters discard: two sequences with the same number of each event can produce different final spins when their event order differs. Train the spin axes, angles, and readout jointly with the task model while constraining every update to remain on the sphere.
Useful8/10
Difficulty5/10
Novelty8/10
✗ Failed on benchmark
2026
Use the generalized Cramér–Rao relation to adapt the inverse-temperature or noise schedule of an energy-based sampler, diffusion sampler, or stochastic optimizer. The controller limits each temperature change according to the measured energy variance and Fisher information, preventing uncontrolled changes in the sampled energy distribution while allowing larger steps in insensitive regions.
Useful8/10
Difficulty4/10
Novelty7/10
✗ Failed on benchmark
2026
Attach one neural value head to each generalized reach-avoid subtask and compose these heads into a critic for sequential or timed temporal-logic goals. The policy is trained to increase the composed value while an auxiliary Hamilton-Jacobi residual trains each local head against the learned or known dynamics. This replaces a single poorly conditioned long-horizon objective with short-horizon certificates whose composition has an explicit logical meaning.
Useful8/10
Difficulty6/10
Novelty7/10
✗ Failed on benchmark
2026
Replace a fixed robust-RL ambiguity radius with a radius computed from the agent’s current belief over environment models. High posterior entropy enlarges the Wasserstein uncertainty set and suppresses catastrophic actions; posterior concentration automatically reduces conservatism and approaches ordinary expected-reward planning.
Useful8/10
Difficulty6/10
Novelty6/10
✗ Failed on benchmark
2026
Train or sample a neural state-space model in trajectory space rather than drawing complete rollouts independently. Construct a space-time path graph whose vertices are latent states and local transition events, then update connected clusters of the entire trajectory using conditional Gibbs or Swendsen-Wang-like moves while preserving fixed initial, terminal, or event-count constraints. This should replace exponentially small forward-rollout success probabilities with local conditional updates…
Useful8/10
Difficulty7/10
Novelty7/10
✗ Failed on benchmark
2026
Split a learned transition model into a contractive nominal branch and a high-capacity excursion branch, and blend them using calibrated epistemic uncertainty. The nominal branch is used exclusively in the well-supported region, while the excursion branch is activated when the current latent state leaves that region, preventing flexible model errors from being recursively amplified during ordinary rollouts.
Useful8/10
Difficulty5/10
Novelty6/10
✗ Failed on benchmark
2026
Train or evaluate a neural dynamical model using many independently restarted finite-precision trajectories instead of one very long rollout. Detect repeated hidden states or quantized state hashes and terminate a segment before its digital transient-plus-period scale, preventing duplicate futures from dominating Lyapunov, loss, and long-horizon forecast estimates.
Useful8/10
Difficulty4/10
Novelty7/10
✗ Failed on benchmark
2026
Replace uniformly spaced history taps in a neural state-space encoder with a fixed or learned set of non-uniform delays. Regularize the resulting delay-observation matrix to have a large smallest singular value, which makes latent-state reconstruction less sensitive to irregular timestamps and observation noise. This is directly applicable to event-based data, missing timestamps, and systems with multiple time scales.
Useful8/10
Difficulty6/10
Novelty7/10
✗ Failed on benchmark
2026
Evaluate a neural network’s learned state by comparing its normal future-task performance with a matched blind counterfactual in which the stored representation, adapter, optimizer state, or memory slots are inaccessible and the model must re-optimize from the same compute budget. Train or select models to maximize this operational value rather than training loss or mutual information with the training data. The method should suppress nuisance memorization because information that cannot…
Useful8/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
When several action branches have nearly equal Q-values, select among them using their long-horizon transition consequences rather than only noisy one-step critic values. Construct a finite sampled approximation to the paper's marked tangential Bellman operator: each candidate receives a local deficit mark and a continuation-value mark, and the branch scores are iterated through a discounted fixed point. Under a perturbation of size comparable to the finite-pool extreme-value gap, the resulting…
Useful8/10
Difficulty7/10
Novelty8/10
✗ Failed on benchmark
2026
Replace a fixed number of randomly sampled continuous actions with a state-dependent candidate pool whose size is chosen from the predicted extreme-value error of the best candidate. If the local action deficit has order \(\|u-u^\star\|^\kappa\) in an effective dimension \(d\), the best sampled action has expected Bellman error proportional to \(N^{-\kappa/d}\). This gives an explicit stopping rule for increasing the pool only when the estimated residual action error is larger than the…
Useful8/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Replace ordinary modality-specific residual fusion with a switched observer whose latent correction depends on the currently available channel. The individual channels are allowed to be insufficient to reconstruct the latent state; stability is enforced over the full switching cycle, so complementary intermittent observations can jointly maintain a stable representation.
Useful8/10
Difficulty6/10
Novelty7/10
✗ Failed on benchmark
2026
Replace a standard softmax-gradient update for a probability vector with a two-stage KL Mirror-Prox update. The predictor evaluates the population-dependent cost at the current distribution, and the corrector evaluates it at the predicted distribution, reducing oscillation when routing or attention costs are coupled across tokens or samples.
Useful8/10
Difficulty5/10
Novelty5/10
✗ Failed on benchmark
2026
Train a neural forecaster or policy network to preserve the pairwise ordering that determines profitable charge and discharge decisions, rather than optimizing only pointwise forecast error. Combine a conventional prediction loss with a pairwise ranking loss weighted by the economic price gap, then pass the prediction through a feasibility-aware storage scheduler.
Useful8/10
Difficulty5/10
Novelty5/10
✗ Failed on benchmark
2026
Constrain the first convolutional layer, or every convolutional layer, by averaging each kernel over the 48 rotations and reflections of the cubic point group. A scalar 3D field then receives exactly the same prediction after any lattice rotation or reflection, eliminating the need to learn equivalent crystallographic orientations from separate examples.
Useful8/10
Difficulty4/10
Novelty5/10
✗ Failed on benchmark
2026
For sparse terminal-reward RL, first solve a discounted version of the fixed-policy Bellman equation to detect states with zero probability of ever reaching success. Clamp these dead states to value zero, retain successful terminal states at value one, and then run undiscounted stochastic policy evaluation and greedy improvement on the reduced state space. The clamp removes the nonunique harmonic solutions that make direct undiscounted TD unstable or initialization-dependent.
Useful8/10
Difficulty5/10
Novelty7/10
✗ Failed on benchmark
2026
Train the raw neural policy to anticipate the safety shield instead of relying on it indefinitely. Add the expected process-correction distance to the PPO objective and adapt its multiplier with a primal-dual update so that policy correction remains below a specified budget.
Useful8/10
Difficulty5/10
Novelty6/10
✗ Failed on benchmark
2026
Use discovered infinitesimal generators to create small continuous transformations of hidden states and force a neural predictor to commute with those transformations. This converts symmetry discovery into self-supervised augmentation without prespecifying a group, canonical coordinates, or hand-designed equivariant layers.
Useful8/10
Difficulty5/10
Novelty6/10
✗ Failed on benchmark
2026
When candidate rewards are noisy, do not score every candidate once and immediately retrain on the apparent elites. Sequentially resample only candidates near the selection boundary and certify their ordering with a mean, median-of-means, trimmed mean, or sign test selected according to the observed tail behavior.
Useful8/10
Difficulty4/10
Novelty5/10