PyTorchKR Papers

최신 논문

총 278편
태그·기간 등 상세 필터

카테고리 탐색

arXiv:2605.26373v2cs.LG

Online Learning on Hidden-Convex Losses via Algorithmic Equivalence: Optimal Regret, Geometric Barrier, and Bandit Feedback

We study adversarial online learning with hidden-convex losses, i.e., nonconvex losses that become convex after a nonlinear reparameterization. Ghai, Lu and Hazan (2022) proved that, under geometric and smoothness assumptions, online gradient descent (OGD) on such nonconvex losses approximately simulates online mirror descent (OMD) on the underlying convex losses with a suitable regularizer, yielding $\mathcal{O}(T^{2/3})$ regret. They left open whether the optimal $\Theta(\sqrt{T})$ regret from online convex optimization can be recovered in this hidden-convex setting. We answer this question affirmatively. More specifically, via a sharper discrete-time algorithmic equivalence argument, we prove that OGD achieves $\mathcal{O}(\sqrt{T})$ regret under the same assumptions, matching the optimal worst-case rate for adversarial online convex optimization. We also address another open question of Ghai, Lu and Hazan (2022) by clarifying the geometry required for this algorithmic equivalence. We replace the diagonal-Jacobian sufficient condition with a necessary-and-sufficient Hessian compatibility condition, thereby expanding the class of admissible reparameterizations. We complement our tight regret bound with a lower bound showing that OGD suffers $\Omega(T)$ regret even with a stepsize of order $1/\sqrt{T}$, for a smooth Hessian incompatible reparameterization between convex domains and an adversarial sequence of hidden-convex losses. Finally, we extend our analysis to one-point bandit feedback and prove a $\mathcal{O}(T^{3/4})$ expected regret bound for bandit OGD with spherical smoothing, matching its classical rate on convex losses.

arXiv:2605.11289v2cs.LG

Quotient-Categorical Representations for Bellman-Compatible Average-Reward Distributional Reinforcement Learning

Average-reward reinforcement learning requires estimating the gain and a state-dependent bias, which solves the Poisson equation only up to an additive constant. Distributional analogues of the bias inherit this same ambiguity. We introduce a quotient-space formulation in which state-indexed laws are identified up to a common translation, together with a categorical parameterization that respects this symmetry. On this space, we define a projected average-reward distributional operator and show that it is well-defined, non-expansive in a coordinate Cram\'er metric, and admits fixed points. We interpret the projected fixed points through geometrically-stopped differential returns, whose categorical operators are contractive and whose fixed points converge in the Abel sense to a canonical fixed point of the projected undiscounted operator. We show that the means of the projected fixed points match the scalar bias up to an error caused by endpoint clipping of the categorical projections, and we bound this error. Finally, we establish almost-sure convergence and finite-iteration residual bounds for temporal-difference recursions with gain-centered rewards under i.i.d. and Markovian sampling. When the gain is unknown, we couple the recursion with an online gain estimator and prove the same guarantees under Markovian sampling, with a bound on the gain error.

arXiv:2605.10219v3math.OC

Parameterized Complexity of Stationarity Testing for Piecewise-Affine Functions and Shallow CNN Losses

We study the parameterized complexity of testing approximate first-order stationarity at a prescribed point for continuous piecewise-affine (PA) functions, a basic task in nonsmooth optimization. PA functions form a canonical model for nonsmooth stationarity testing and capture the local polyhedral geometry that appears in ReLU-type training losses. Recent work of Tian and So (2025) shows that testing approximate stationarity notions for PA functions is computationally intractable in the worst case, and identifies fixed-dimensional tractability as an open direction. We address this direction from the viewpoint of parameterized complexity, with the ambient dimension $d$ as the parameter. In this paper, we give XP algorithms in fixed dimension for the tractable sides, and prove W[1]-hardness for the complementary sides. Moreover, lower bounds under the parameterized $k$-Clique hypothesis rule out deterministic algorithms running in time $\rho(d)s^{o(d)}$ for any computable function $\rho$, where $s$ denotes the total binary encoding length of the instance. As a further consequence, our results yield the corresponding parameterized complexity picture for testing local minimality of continuous PA functions. We also extend our hardness results to a family of shallow ReLU CNN training losses, tested in the trainable weight space. Thus, the same parameterized-complexity picture also appears for simple CNN training losses. Notably, our lower bounds assume that $k$-Clique on $N$-vertex graphs admits no deterministic $\rho(k)N^{o(k)}$-time algorithm for any computable function $\rho$. The recent refutations of the Exponential Time Hypothesis (ETH) (OpenAI, 2026) and the stronger fine-grained $k$-Clique hypothesis (Vassilevska Williams and Zamir, 2026) do not contradict this assumption.

arXiv:2602.12517v2cs.LG

Bench-MFG: A Benchmark Suite for Learning in Stationary Mean Field Games

The rise of Mean Field Games (MFGs) has fostered a growing family of algorithms designed to solve large-scale multi-agent systems. However, the field currently lacks a standardized evaluation protocol, forcing researchers to rely on bespoke, isolated, and often simplistic environments. This fragmentation makes it difficult to assess the robustness, generalization, and failure modes of emerging methods. To address this gap, we propose a comprehensive benchmark suite for MFGs, focusing on the discrete-time, discrete-space, stationary setting. We introduce a taxonomy of problem classes, ranging from no-interaction and monotone games to potential and dynamics-coupled games, and provide prototypical environments for each. Furthermore, we present MF-Garnets, a method for generating random MFG instances to facilitate rigorous statistical testing. We benchmark a variety of learning algorithms across these environments, including a novel population-based approach (MF-PSO). To our knowledge, this is the first benchmark for MFGs.

arXiv:2602.03461v2cs.LG

Soft-Radial Projection for Constrained End-to-End Learning

Constrained end-to-end learning trains neural networks whose outputs must satisfy feasibility constraints, such as resource, budget, or operational limits. While projection and optimization layers guarantee constraint satisfaction, boundary-based projections can introduce unfavorable optimization geometry: regions of the unconstrained output space are mapped to the same active face of the feasible set. This makes the resulting map locally rank-deficient, suppressing gradients in constrained directions and potentially impairing optimization - a bottleneck known as gradient saturation. We propose Soft-Radial Projection (SRP), a differentiable reparameterization layer that maps unconstrained network outputs into the relative interior of a convex feasible set by smoothly rescaling rays from a strictly feasible anchor point. Unlike hard radial or orthogonal projections, the resulting map is one-to-one and has a full-rank Jacobian almost everywhere, while guaranteeing strict feasibility. We prove that constrained networks equipped with SRP retain universal approximation guarantees. For polytopes and balls, the layer admits a closed-form forward pass, while for general convex sets, the ray-boundary intersection reduces to a one-dimensional search. Experiments on decision-focused learning benchmarks demonstrate strong solution quality at low computational cost, with particularly favorable results in portfolio optimization and substantially lower inference latency than differentiable optimization layers in the evaluated settings.

arXiv:2601.20983v2cs.LG

Monotone Optimisation with Learned Projections

Optimisation algorithms typically assume oracle access to objectives and constraints, yet in many applications these functions are available only through data. To recover them, a common approach learns a surrogate for each function and optimises over it, but this can be unnecessarily expensive and may fail to preserve the structure required by the optimisation algorithm. We study this problem for monotone optimisation, a class of non-convex problems which can be solved globally using the Polyblock Outer Approximation (POA) algorithm. Our key observation is that POA does not require the constraint functions themselves, but only a scalar quantity we call the radial inverse. Learning this quantity directly replaces POA's inner bisection loop over a learned constraint with a single radial inverse evaluation. We characterise exactly which maps arise as radial inverses of monotone constraints, and relate their approximation error to perturbations of the feasible set and optimal objective value. Building on these results, we introduce Homogeneous-Monotone Radial Inverse (HM-RI) networks, structured min-max models that enforce the monotonicity and positive-homogeneity of radial inverses by construction. Across synthetic monotone optimisation problems and a cellular transmit-power application, HM-RI improves solution quality over direct constraint surrogates and local optimisation baselines by up to 58.6 percentage points relative to the Oracle reference objective, while achieving a 3.0-4.7x speed-up over bisection-based POA with learned constraint surrogates, with the largest solution-quality gains occurring in data-limited and piecewise-constant settings.

arXiv:2512.17473v4eess.SP

Alternating Direction Method of Multipliers for Nonlinear Matrix Decompositions

We present an algorithm based on the alternating direction method of multipliers (ADMM) for solving nonlinear matrix decompositions (NMD). Given an input matrix $X \in \mathbb{R}^{m \times n}$ and a factorization rank $r \ll \min(m, n)$, NMD seeks matrices $W \in \mathbb{R}^{m \times r}$ and $H \in \mathbb{R}^{r \times n}$ such that $X \approx f(WH)$, where $f$ is an element-wise nonlinear function. We evaluate our method on several representative nonlinear models: the rectified linear unit activation $f(x) = \max(0, x)$, suitable for nonnegative sparse data approximation, the component-wise square $f(x) = x^2$, applicable to probabilistic circuit representation, and the MinMax transform $f(x) = \min(b, \max(a, x))$, relevant for recommender systems. The proposed framework flexibly supports diverse loss functions, including least squares, $\ell_1$ norm, and the Kullback-Leibler divergence, and can be readily extended to other nonlinearities and metrics. We illustrate the applicability, efficiency, and adaptability of the approach on real-world datasets, highlighting its potential for a broad range of applications.

arXiv:2510.09034v3cs.LG

Convergence of optimizers implies eigenvalues filtering at equilibrium

Ample empirical evidence in deep neural network training suggests that a variety of optimizers tend to find nearly global optima. In this article, we adopt the reversed perspective that convergence to an arbitrary point is assumed rather than proven, focusing on the consequences of this assumption. From this viewpoint, in line with recent advances on the edge-of-stability phenomenon, we argue that different optimizers effectively act as eigenvalue filters determined by their hyperparameters. Specifically, the standard gradient descent method inherently avoids the sharpest minima, whereas Sharpness-Aware Minimization (SAM) algorithms go even further by actively favoring wider basins. Inspired by these insights, we propose two novel algorithms that exhibit enhanced eigenvalue filtering, effectively promoting wider minima. Our theoretical analysis leverages a generalized Hadamard--Perron stable manifold theorem and applies to general semialgebraic $C^2$ functions, without requiring additional non-degeneracy conditions or global Lipschitz bound assumptions. We support our conclusions with numerical experiments on feed-forward neural networks.

arXiv:2510.00569v2stat.ML

Guaranteed Noisy CP Tensor Recovery via Riemannian Optimization on the Segre Manifold

Recovering a low-CP-rank tensor from noisy linear measurements is a central challenge in high-dimensional data analysis, with applications spanning tensor PCA, scalar-on-tensor regression, tensor-on-tensor regression, and beyond. We exploit the intrinsic geometry of rank-one tensors by casting the recovery task as an optimization problem over the Segre manifold, the smooth Riemannian manifold of rank-one tensors. This geometric viewpoint yields two powerful algorithms: Riemannian Gradient Descent (RGD) and Riemannian Gauss-Newton (RGN), each of which preserves feasibility at every iteration. Under mild noise assumptions, we prove that RGD converges at a local linear rate, while RGN exhibits an initial local quadratic phase followed by linear refinement to the statistical error. Extensive synthetic experiments validate these convergence guarantees and demonstrate the practical effectiveness of our methods.

arXiv:2509.20919v2cs.GT

Efficient Kernelized Learning in Polyhedral Games Beyond Full-Information: From Colonel Blotto to Congestion Games

We examine the problem of efficiently learning coarse correlated equilibria (CCE) in polyhedral games, that is, normal-form games with an exponentially large number of actions per player and an underlying combinatorial structure. Prominent examples of such games are the classical Colonel Blotto and congestion games. To achieve computational efficiency, the learning algorithms must exhibit regret and per-iteration complexity that scale polylogarithmically in the size of the players' action sets. This challenge has recently been addressed in the full-information setting, primarily through the use of kernelization. However, in the case of the realistic, but mathematically challenging, partial-information setting, existing approaches result in suboptimal and impractical runtime complexity to learn CCE. We tackle this limitation by building a framework based on the kernelization paradigm. We apply this framework to prominent examples of polyhedral games -- namely the Colonel Blotto, graphic matroid and network congestion games -- and provide computationally efficient payoff-based learning algorithms, which significantly improve upon prior works in terms of the runtime for learning CCE in these settings.

arXiv:2505.01353v3math.OC공식 코드 · Papers with Code

Differentiable Nonlinear Model Predictive Control

The efficient computation of parametric solution sensitivities is a key challenge in the integration of learning-enhanced methods with nonlinear model predictive control, as their availability is crucial for many learning algorithms. This paper discusses the computation of solution sensitivities of general nonlinear programs using the implicit function theorem and smoothed optimality conditions treated in interior-point methods (IPM). We introduce sequential barrier quadratic programming (SBQP), which generalizes SQP with an IPM quadratic programming (QP) solver by allowing the barrier parameter to approach a prescribed nonnegative target value, recovering classical SQP when this target is zero. Previous works are limited to convex or unconstrained formulations, or lack an implementation for efficient sensitivity evaluation. The publication is accompanied by an efficient open-source implementation within the `acados` framework, providing both forward and adjoint sensitivities for general optimal control problems, achieving speedups exceeding 3x over the state-of-the-art solvers `mpc.pytorch` and `cvxpygen`.

arXiv:2611.09887v1cs.LG

On the (non-)closedness of the set of ReLU networks

We prove that the set $\mathcal{N}_N$ of ReLU networks with one hidden layer and at most $N\in\mathbb N$ neurons is weakly closed in $W^{1,p}(\Omega)$, for a bounded connected open set $\Omega\subset\mathbb R^d$. Consequently, every coercive, weakly lower semicontinuous energy on this space attains its minimum over $\mathcal{N}_N$. This covers the Deep Ritz method with convex integrands, Sobolev training and Lipschitz learning. Moreover, as the weight decay parameter tends to zero, minimizers of the regularized problem converge, up to subsequences, to minimizers with minimal parameter norm. We also prove that the set of ReLU networks with two hidden layers of bounded widths is not closed in $W^{1,p}(\Omega)$ or in $C(\overline\Omega)$. This disproves a conjecture of Petersen, Raslan and Voigtlaender.

arXiv:2611.09778v1cs.LG

Quasipolynomial-Time Planning for Decodable POMDPs

We study planning for finite-horizon partially observable Markov decision processes (POMDPs) with finite state, action, and observation spaces. Both exact and approximate planning are known to be computationally hard in this setting, so prior work has explored natural assumptions that make planning easier. Recent work [Golowich et al., 2022] gave quasipolynomial-time approximate planning algorithms for observable POMDPs. These are POMDPs whose observation channels satisfy an $\ell_1$-observability condition, but notably, no assumptions are required on the transition dynamics. In a similar vein, for every fixed $\rho>0$, we give quasipolynomial-time approximate planning algorithms for $\rho$-decodable POMDPs. We show that decodability is implied by both $\ell_1$- and $\ell_2$-observability and is strictly weaker than observability. Moreover, we show that the optimal decodability constant can be computed in polynomial time, whereas computing the optimal $\ell_1$-observability constant is NP-hard in the worst case. We also show that the fast filter-forgetting guarantee underlying the approach of [Golowich et al., 2022] does not extend directly from $\ell_1$-observability to constant $\ell_2$-observability. Thus, decodability is an easy-to-check, unifying assumption under which quasipolynomial-time approximate planning for POMDPs is possible.

arXiv:2611.09723v1cs.LG

Fixed-point acceleration on Neural Wasserstein manifolds

We propose a fixed-point acceleration method for optimization over probability measures, applicable to a variety of machine learning problems. We represent probability measures as pushforwards of a reference distribution through a neural network map and equip the parameter space with the induced pullback Wasserstein metric. This geometric structure allows natural gradient descent to be interpreted as a fixed-point iteration on a Riemannian manifold. We accelerate this iteration using Riemannian Anderson mixing method, which exploits historical residuals to capture approximate second-order information without forming Hessians. In the present parametric setting, the manifold is represented by Euclidean coordinates endowed with a point-dependent metric, simplifying retraction and vector transport. We evaluate the method on multimodal sampling, PDE-based inverse problems, aggregation dynamics, Bayesian logistic regression, and generative modeling. Across the tested tasks, Anderson mixing reduces the number of natural-gradient iterations and achieves competitive performance relative to SGD and Adam in terms of both loss-gradient evaluations and wall-clock time. These results suggest that fixed-point acceleration is an efficient tool for improving the computational efficiency of training on parametric Wasserstein methods.

arXiv:2611.09560v1cs.LG

Solver-Free Decision-Focused Learning via Neural Optimization Layers for Edge Selection

Decision-focused learning trains prediction models to improve downstream decisions, but many methods require costly solver calls or optimal-solution labels. We propose a solver-free decision-focused learning framework for NP-hard edge-selection problems with uncertain parameters in the objective, the constraints, or both. A neural optimization layer is first trained without solution labels, using problem objectives and constraint penalties. Graph-informed edge scoring and iterative refinement encourage concentrated and structurally consistent continuous decisions. The layer is then frozen, and a prediction model is trained through its continuous output using a problem-specific loss evaluated under observed parameters. At deployment, the predictor, optimization layer, and a lightweight decoder produce decisions without solving the original NP-hard problem. Experiments on the asymmetric traveling salesperson and endpoint-free orienteering problems show consistently better decision quality than prediction-focused learning and improvements over most evaluated decision-focused baselines. Our framework requires less training time than the evaluated decision-focused baselines, even when optimization-layer pretraining is included. Neural deployment also produces decisions faster than solver-based deployment, offering a better quality-efficiency trade-off. The layer also reduces training cost in existing decision-focused methods when used as an approximate optimization oracle.

arXiv:2611.09217v1cs.LG

Provable State Estimation with Recurrent Models

Estimating the state of a dynamical system from observations is a crucial task in robotics and control. Despite their empirical success, recurrent models often lack formal guarantees on convergence and robustness. We bridge this gap by analyzing these architectures from the perspective of nonlinear observer theory. By framing RNNs as a time-discretization of continuous latent dynamical systems (Neural ODEs), we demonstrate that enforcing a contraction condition on the latent flow guarantees the existence of a smooth, injective map of the true physical state into the latent space. We prove that this architecture acts as a tunable observer: by increasing a single scalar gain that scales the latent vector field, the estimation error can be driven to an arbitrarily small neighborhood of zero at an exponentially fast rate. We further establish quantitative bounds on the MLP complexity required to decode this latent state, and extend our formal guarantees to the discrete-time residual RNN implementation. We validate the theory on a real mobile robotic task, where a provably contracting model outperforms a vanilla GRU with substantially more stable training. We also report that unconstrained GRUs converge to contracting dynamics on this task, suggesting that contraction is an emergent property of well-trained sequence models.

arXiv:2611.08775v1cs.LG

On Lipschitz Explosion in Deep Neural Networks with Normalization: Consequences for Optimization and Adversarial Robustness

Can a network look stable at every layer and still be exponentially unstable as a whole? The Lipschitz constant of a neural network is a basic stability quantity appearing in optimization guarantees, generalization bounds, and robustness certificates. Deep networks naturally admit two such notions, input Lipschitzness, measuring sensitivity to data perturbations, and parameter Lipschitzness, measuring sensitivity to weight perturbations. We prove that, for deep networks with normalization layers, with batch normalization as the canonical example, both Lipschitz constants can grow exponentially with depth and hence with parameter dimension, even under strong per-layer norm control such as $\|W_\ell\|_2\le 1$, and already for linear activations. Thus, a mechanism widely used to stabilize training can create severe worst-case instabilities that are invisible locally, i.e., from layerwise norm bounds alone. Parameter-Lipschitz explosion turns Lipschitz-dependent nonsmooth optimization guarantees into exponential-in-dimension bounds for deep normalized networks, while input-Lipschitz explosion makes worst-case Lipschitz-based generalization and robustness certificates vacuous. The latter also yields a concrete rank-separation mechanism for adversarial vulnerability. The theory predicts that perturbations introducing new singular directions should be amplified much more strongly than perturbations at the same noise level that remain within the input's original singular subspace. Experiments on MNIST, Fashion-MNIST, and CIFAR-10 support this prediction, showing that rank-creating perturbations cause substantially sharper drops in accuracy, confidence, and margins than same-subspace perturbations.

arXiv:2611.08505v1stat.ML

Drift-admissibility via Optimal Transport

We introduce structure-aware optimal transport diagnostics for assessing whether empirical time-series dynamics are compatible with a prescribed predictable-drift class. Our approach translates a pathwise property of stochastic processes into a finite-dimensional projection problem over empirical transition laws, encoding admissibility as row-wise moment constraints on local couplings and quantifying deviations via a Kullback--Leibler projection gap $G_{\lambda,k}$. This yields both a principled diagnostic and a scalable Bregman--Dykstra algorithm that retains the efficiency of Sinkhorn-type methods while incorporating structural constraints. Calibrated against a Brownian reference null, $G_{\lambda,k}$ separates process classes that are indistinguishable by second-order statistics alone---including fractional Brownian motion and BSS processes sharing identical Hurst indices---without asymptotic resampling or continuous-time recovery. This positions the method as a general tool for structure-aware model validation of sequential data.

arXiv:2611.04171v2math.OC

Descending Everest, Part II: Oracle Complexity of Minimizing a Sum of Convex $(L_0,L_1)$-Smooth Functions

We study the minimization of the average of $n$ convex $(L_0,L_1)$-smooth functions: the Hessian norm of each function is bounded by an affine function of its gradient norm. We assume known upper bounds $R$ on the distance from the starting point to a minimizer and $\delta$ on the difference between the minimum of the average and the average of the infima of the components. We derive upper and lower bounds on the oracle complexity of reaching accuracy $\varepsilon$ in function value in expectation. Up to logarithmic factors, both bounds are of order $n(1+(L_1R)^{2/3}) +\sqrt{\frac{n(L_0+L_1^2\delta)R^2}{\varepsilon}+n(L_1R)^2}$. The upper bound is attained by the proposed method YETI. The method combines ideas of accelerated proximal methods and variance reduction methods with a heuristic for choosing additional points. The number of oracle calls made by the heuristic is bounded, which preserves the upper bound on the total number of queries needed to reach the target accuracy. In the experiments, YETI is compared with first-order methods on synthetic problems and Poisson regression problems. We also examine the stopping criterion and the effect of the heuristic separately. The lower bound is obtained by an explicit construction of families of instances that are hard for any first-order method.

arXiv:2610.34590v2cs.LG

Sparse Sum-of-Squares Layers for Certified Learning

A central limitation of modern machine learning systems is the lack of formal certificates guaranteeing their behavior beyond empirical, sample-based evaluation. Sum-of-squares (SOS) methods provide such guarantees by certifying polynomial inequalities over continuous domains via semidefinite programming (SDP). While prior work has introduced SDP layers as differentiable components in neural networks, repeated learning calls can lose the computational benefit of the forward program's sparse representation. We propose Sparse SOS Layers, differentiable value layers whose backward passes reuse the local coordinates of the forward solve. Under a fixed template and learned objective coefficients, each layer avoids building dense certificates and reuses its sparse structure across training steps. Compared with existing methods, our approach achieves up to 2.7 times speedup in total warm-call time, with nearly 100 times lower memory usage (1.7 vs. 160 GiB) for relatively large problem sizes for the tested dense parameterization. Experiments cover sparse polynomial certification, pendulum control, combinatorial relaxations, robotics projection, oscillator chains, and pose graph synchronization. A design study achieves zero certified violations across unseen problem instances and sizes, while requiring roughly 7 times fewer optimization solves than estimating gradients by perturbing the objective coefficients and re-solving the problem (finite differences). These results show that sparse certificate values can be effectively integrated into learning pipelines when the backward pass respects the verifier's structure.