ML: Optimizer

Machine-learning ideas tagged Optimizer in the ML taxonomy of the Math2NN corpus.

728 ideas found

Unverified 2026

L-Stable Trajectory-Derivative Optimizer

Replace an explicit gradient step by an implicit correction using the trajectory derivative \(Dg(\theta)g(\theta)=H(\theta)g(\theta)\), where \(g=\nabla f\) and \(H=\nabla^2 f\). The update should strongly damp high-curvature or stiff modes while preserving fourth-order matching of the local linearized dynamics. Start with a self-contained fourth-order L-stable rational prototype, then compare it with the paper's exact two-stage coefficients after recovering those coefficients from the full…

Useful6/10
Difficulty7/10
Novelty7/10
Paper: An L-Stable Sequential Two-Stage Fourth-Order Method with ADER Trajectory Derivatives for Stiff Transport--Relaxation Systems arXiv:2608.03256
Unverified 2026

Asymptotic-Preserving Adjoint for Stiff Relaxation Layers

Replace ordinary reverse-mode differentiation through a long sequence of stiff relaxation updates with a projected adjoint that separates slow conserved features from rapidly relaxing residual features. The neural layer can use large outer time steps even when its internal relaxation time is very small, while reconstructing only the microscopic gradient component required by the preceding layer.

Useful6/10
Difficulty6/10
Novelty7/10
Paper: An asymptotic-preserving adjoint unified gas kinetic scheme for sensitivity analysis arXiv:2608.03236
Unverified 2026

Parameter-Free Certified Augmented-Lagrangian Fine-Tuning

Replace a manually tuned penalty optimizer with an inexact augmented-Lagrangian optimizer for neural parameters subject to exact linear constraints such as parameter tying, zero-sum filters, conservation constraints, or structured adapter constraints. Each outer iteration approximately minimizes the augmented Lagrangian using an accelerated proximal-gradient inner loop, and stops when an explicitly computed stationarity certificate reaches a target determined from the current feasibility…

Useful6/10
Difficulty5/10
Novelty7/10
Paper: Optimal Nonergodic Primal-Dual Complexity of Efficient Inexact Parameter-Free Augmented Lagrangian Methods arXiv:2608.03170
Unverified 2026

Volterra-Fredholm delay-compensated optimizer

Replace a delayed optimizer state or gradient by a causal lower-triangular history transformation that predicts the current descent direction from recently stored states and inputs. Use Fredholm terms to incorporate the recent history and Volterra terms to preserve causal invertibility, then apply the optimizer update in transformed coordinates. This targets oscillation and divergence caused by concurrent delays in distributed or asynchronous training.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: Stabilization of First-Order Partial Integro-Differential Equations with Concurrent Input and State Delays arXiv:2608.02851
Unverified 2026

Multifractal Noise-Stability Monitor

Monitor moments of the network's response to independent stochastic forward passes instead of tracking only mean loss or mean activation variance. Nonlinear moment scaling detects intermittent and heterogeneous sensitivity, allowing a controller to reduce noise or learning rate before average metrics reveal instability.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: Universal crossovers in weakly-monitored quantum critical states arXiv:2608.02716
Unverified 2026

Performance-Gated Adaptation Freeze

Add a low-cost performance monitor to an online-adapted neural network and freeze gradient updates after the monitored error has stayed below a target for a dwell interval. The gate prevents continued low-information updates, which otherwise cause parameter drift under weak excitation, noisy observations, or stationary data. Hysteresis allows adaptation to restart after a genuine performance deterioration.

Useful6/10
Difficulty3/10
Novelty6/10
Paper: Performance-based Adaptation Termination for Preventing Parameter Drift in Adaptive Vibration Suppression arXiv:2608.02570
Unverified 2026

Nonadaptive multiscale one-bit gradient sketch

Replace communicated floating-point gradients in synchronous federated or data-parallel training with one-bit threshold queries whose thresholds are sampled publicly before gradients are observed. Use several fixed geometric amplitude scales so the same protocol handles unknown gradient means and heavy-tailed client updates without an interactive localization round. Decode each coordinate from the scale whose neighboring estimates are statistically consistent, then apply the decoded aggregate…

Useful6/10
Difficulty6/10
Novelty6/10
Paper: Interaction Is Not Necessary for Order-Optimal 1-Bit Mean Estimation arXiv:2608.02538
Unverified 2026

Uniform-History Reset Optimizer

Augment gradient descent with stochastic relocations to uniformly sampled historical parameter vectors. In expectation, the optimizer receives a non-Markovian correction toward the running average of all previous iterates, which can suppress runaway directions and revisit earlier basins instead of remaining trapped in a sharp or unstable region.

Useful6/10
Difficulty4/10
Novelty7/10
Paper: Quantum resetting with memory arXiv:2608.02297
Unverified 2026

Exponential Frequency-Map Optimizer Monitor

Estimate persistent frequencies in a neural-network training trajectory using a smooth weighted Birkhoff average instead of a rectangular moving average. Use the estimated frequency vector to detect low-order resonances between optimizer oscillations, gradient-noise cycles, and validation-loss oscillations, then trigger a learning-rate or momentum intervention before divergence.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: Laskar's frequency map analysis revisited arXiv:2608.02182
Unverified 2026

Random-start learned AMP

Construct an unrolled phase-retrieval network that begins with an isotropic Gaussian estimate rather than a spectral initializer. Retain the AMP residual correction and Onsager subtraction, but learn the scalar measurement denoisers and step sizes; use several random starts and select the iterate with the lowest measurement residual.

Useful6/10
Difficulty5/10
Novelty4/10
Paper: Approximate Message Passing with Random Initialization for Phase Retrieval arXiv:2608.01654
Unverified 2026

Collision-free energy-kernel prototype flow

Replace ordinary codebook or prototype updates with the paper's diagonal-free particle dynamics in feature space. Each prototype is attracted toward the current minibatch feature distribution, while pairwise negative-distance interactions spread prototypes apart and suppress codebook collapse. The exponent q in (0,2) controls the interaction profile, with q less than 1 producing a stronger near-collision force but requiring numerical regularization.

Useful6/10
Difficulty4/10
Novelty6/10
Paper: Wasserstein gradient flows of Maximum Mean Discrepancy with energy kernels arXiv:2608.01182
Unverified 2026

Orientation-Preserving Simplex Deformation Layer

Represent a neural deformation of a mesh or simplicial graph by vertex positions \(f\), and constrain every oriented simplex to retain positive signed volume. Add a logarithmic barrier during feasible optimization and use a feasibility-restoration phase for initially inverted elements, turning foldover prevention into a hard geometric invariant rather than a soft penalty.

Useful6/10
Difficulty6/10
Novelty7/10
Paper: A Novel Bijective Angle and Volume-preservation Balanced Parameterization for $n$-dimensional Manifolds arXiv:2608.01073
Unverified 2026

Benign-Misfit Large-Step Phase

Add a deliberate large-constant-learning-rate phase in which training loss is not forced monotonically toward interpolation. The phase is intended to calibrate shared, high-signal directions before the optimizer memorizes example-specific nuisance directions, and should be stopped when validation error is minimized even if training error remains high.

Useful6/10
Difficulty5/10
Novelty6/10
Paper: The Fourth Quadrant: A Stylized View of Benign Misfitting arXiv:2608.01032
Unverified 2026

Prototype Distance-Field Safety Layer

Store a finite library of successful robot configurations or action-conditioned waypoints and construct a smooth soft minimum of their distances. Use the negative distance gradient as a structured action prior, add a learned residual policy, and pass the combined action through a quadratic-program safety layer. This gives a neural controller an explicit attraction basin toward demonstrated solutions while preventing violations of known state constraints.

Useful6/10
Difficulty5/10
Novelty6/10
Paper: Grasp Execution Without a Planner: Configuration-Space Grasp Distance Fields with Certified Safety & Guaranteed Quality arXiv:2608.00600
Unverified 2026

Sparse Learnable Power-Law Head

Attach a symbolic sparse head to a neural encoder instead of using a dense final MLP. The head evaluates a library of learnable power-law and interaction terms on nonnegative learned features, jointly optimizes linear coefficients and exponents, and removes inactive terms with coefficient sparsity. This should provide a compact model with better relative-error behavior on positive targets spanning several orders of magnitude.

Useful6/10
Difficulty5/10
Novelty7/10
Paper: Discovering Explicit Magnetic Core Loss Equations via Learnable Symbolic Sparse Identification arXiv:2608.00379
Unverified 2026

Entropy-Volume Growth Regularization

Model stochastic training or recurrent inference as a random dynamical system and penalize the exponential growth of volumes transported by its Jacobian. This converts the paper's entropy and volume-growth relation into a computable regularizer that discourages chaotic sensitivity while retaining directions needed for fitting.

Useful6/10
Difficulty5/10
Novelty6/10
Paper: Continuity of measure-theoretic entropy for stochastic differential equations arXiv:2608.00370
Unverified 2026

Rational Jacobi Curvature Preconditioner

Replace an ordinary dense or floating-point eigendecomposition of small Hessian or Fisher blocks with a sequence of rational Jacobi rotations. The rotations preserve Euclidean norms and can be stored using fixed-point coefficients, while approximately diagonalizing curvature so the optimizer can use separate coordinate-wise step sizes. This is especially relevant to low-precision training and blocks with mixed-sign curvature.

Useful6/10
Difficulty6/10
Novelty6/10
Paper: Rational Jacobi Rotations and the Complexity of Approximating Mixed Integer Quadratic Programming arXiv:2607.29386
Unverified 2026

Induced-pressure controller for marginal recurrent dynamics

Replace a single-step spectral-radius diagnostic in a recurrent network with a multiscale induced pressure computed from return trajectories. Separate return branches whose Jacobian products remain close to the limiting dynamics from transverse branches that create rapid growth in trajectory complexity, then reduce recurrent gain or optimizer step size when the transverse pressure exhibits the predicted square-root rise near a neutral bifurcation.

Useful6/10
Difficulty6/10
Novelty7/10
Paper: Precise asymptotics at the tip of the Mandelbrot set arXiv:2607.29326
Unverified 2026

Fleming-Viot Stable-Trajectory Optimizer

Run multiple neural-network parameter trajectories in parallel and define divergence, NaNs, loss explosions, or trust-region violations as absorbing failure events. Whenever one replica fails, replace it with a copy of a uniformly selected survivor while tracking the time since its last replacement. This creates an empirical quasi-stationary distribution of robust training states instead of relying on one potentially unstable trajectory.

Useful6/10
Difficulty5/10
Novelty8/10
Paper: Fleming-Viot Selection of the Yaglom Limit for Age-Structured Bellman-Harris Processes, with Application to Livestock Epidemic Surveillance arXiv:2607.29251
Unverified 2026

Moment-Controlled Masked Leader Search

Use a small population of neural parameter vectors and replace isotropic random perturbations with the paper's masked affine move toward the current best candidate. Select the mask probability and migration distance from the closed-form expected step-length and active-dimensionality formulas, allowing large exploratory moves early and progressively focused moves later.

Useful6/10
Difficulty4/10
Novelty7/10
Paper: Linear Proposal Operators and Stochastic Search Geometry in SOMA and Differential Evolution arXiv:2607.29228
Unverified 2026

Oja Gradient-Subspace Optimizer

Track the dominant rank-r subspace of the gradient covariance online, then use that basis to construct a low-rank adaptive update or a controlled preconditioner. Unlike offline PCA refreshes, the Oja flow continuously follows changing training geometry while preserving orthonormality, potentially reducing the cost of second-order or Shampoo-like methods.

Useful6/10
Difficulty5/10
Novelty6/10
Paper: On the Oja-Flow-Based Low-Rank Approximation of Kalman-Bucy Filters for Linear Time-Varying Systems arXiv:2607.29034
Unverified 2026

Interacting Hypothesis-Bank Optimizer

Replace one potentially misinitialized training trajectory with K parallel parameter hypotheses, each representing a different basin or latent explanation, and combine them using loss-derived mode probabilities. Before each update, mix the hypotheses through a transition matrix so that a temporarily poor or incorrect mode can inherit information from a promising mode while retaining multimodal diversity. This is most appropriate for nonconvex networks, latent-variable models, or long-horizon…

Useful6/10
Difficulty6/10
Novelty6/10
Paper: Adaptive Attitude Estimation for Multiple-Surface Object Using Light Curve Glints arXiv:2607.28912
Unverified 2026

Geometric parallel optimizer portfolio

Run several independent copies of an optimizer with geometrically increasing training budgets so that at least one copy is close to the unknown budget required by the problem. Allocate workers using staggered schedules rather than launching a full hyperparameter sweep, and periodically select or promote the best checkpoint using a held-out minibatch stream. This targets training regimes where the required number of updates is unknown because smoothness, gradient variance, or noise level is…

Useful6/10
Difficulty5/10
Novelty5/10
Paper: Adaptivity via a Parallel Architecture for Stochastic Gradient Methods arXiv:2607.28902
Unverified 2026

Stochastic Bilevel Structure Learning

Replace a hard, separately precomputed graph partition or expert assignment with partition-inducing parameters sampled from a learnable Gaussian distribution. Train the neural representation in an inner loop and update the distribution parameters using an outer validation loss, allowing the discovered structure and predictor to co-adapt while retaining gradients through otherwise discrete assignments.

Useful6/10
Difficulty6/10
Novelty5/10
Paper: MPP-GNN: Subject-Adaptive Community Detection for fMRI-Based Alzheimer's Disease Classification arXiv:2607.28681