arXiv subjects

Adam M. Oberman

Publications and source records attributed to Adam M. Oberman.

At least 19 recordsLinked to original sources

arXiv2026

Individual Disempowerment through an Advice Channel: Control Loss when Influence is Endogenous

An AI that can only give advice seems safe: the human is always free to ignore it. That is the premise of the boxing tradition in AI safety, and its long-suspected weak point is that the human who reads the answers is part of the system. We make the fraction $\varepsilon_t$ of behavior that follows the advice a state of a Markov decision process, moved by the advisor's own messages, so that use deepens reliance. Granted a channel rich enough to echo any action the human could take, higher $\varepsilon_t$ weakly lowers every monotone measure of the power of a human with a message-independent fallback. An oracle rewarded by per-round approval cultivates reliance beyond a closed-form patience threshold, so the same reward weights leave the optimal oracle answering in episodic deployments and cultivating in long-memory ones. An influence bound certified once at deployment is blind to that horizon and bounds the loss no lower than its trivial ceiling. An exogenous cap on influence bounds the guarantee the human loses, and a short enough memory reset removes the incentive to cultivate, while neither recovers the value already steered away. In a closed-form example the optimal oracle never cultivates in fifteen-round sessions and does in sixteen.

cs.AI↗

arXiv2026

Fast Rates for Semi-Supervised Learning via Data-Augmentation Graph Regularization

Self-supervised learning matches supervised accuracy from a fraction of the labels, but the labeled-sample efficiency behind this has lacked a theoretical explanation. We provide one. Data augmentation induces a similarity graph on the unlabeled data, so downstream learning on that graph is graph-Laplacian-regularized learning. We prove a fast transductive rate, $O(1/n_L)$ in the number of labels, in place of the supervised $O(1/\sqrt{n_L})$, by carrying the leave-one-out stability apparatus of Johnson and Zhang (JMLR 2007) over to the augmentation graph, and without the unrealistic assumptions of limit-based analyses (exact kernel, generalizing features). The bound makes augmentation quality explicit: the expected error is at most $C/n_L + R_{\mathrm{DA}}(y)$, where the data-augmentation alignment error $R_{\mathrm{DA}}(y)$ is proportional to the graph-cut mass of augmentations that cross a label boundary, so good augmentations let few labels suffice. The analysis uses a streamlined loss that drops the projector, negative-sample, and orthogonality overhead of standard objectives yet still recovers the top-$K$ ideal features in the infinite-data limit, the augmentation-kernel eigenspace studied by Zhai et al. The bound gives a mechanistic account of the accuracy-versus-label-count curve through augmentation quality, verified in a controlled model where the constants are known.

cs.LG↗

arXiv2026

Avoiding unsafe sets when training with Langevin Dynamics

Training a model with noisy gradient descent can be idealized as overdamped Langevin dynamics, and a natural safety question is to bound the probability $\nu_t(\mathcal{A}_H) = \mathbb{P}(Q_t \in \mathcal{A}_H)$ that the trajectory lies in a designated failure region $\mathcal{A}_H$. We study this for a smooth, strongly convex loss in $d$ dimensions, with $\mathcal{A}_H$ separated from the minimizer by an energy gap. At the end of training, the equilibrium mass $\pi(\mathcal{A}_H)$ is exponentially small in $d$, with a complementary energy-barrier rate when the noise is small. Along the trajectory, a shape-free bound $\nu_t(\mathcal{A}_H) \le \pi(\mathcal{A}_H)(1 + \sqrt{\chi_0^2/\pi(\mathcal{A}_H)}\,e^{-mt})$ shows the in-set probability relaxes to (twice) the static value after a burn-in of order $d$, using only the global spectral gap $m$. A worked Ornstein-Uhlenbeck example shows this burn-in is necessary: an angular slice of the equilibrium shell can transiently swell by a factor exponential in $d$, though its equilibrium mass is tiny. To rule this out we introduce a local relaxation rate, defined through the spectral measure of the region's centered indicator rather than a Dirichlet-form Rayleigh quotient. For geometrically isolated regions this rate exceeds the global one, shrinking the burn-in, and with a maximum-principle ceiling it caps the trajectory probability uniformly in time. Strong convexity sets how fast training relaxes, but the shape of the unsafe set decides whether the trajectory bulges through it on the way to equilibrium.

cs.LG↗

arXiv2019

No-collision Transportation Maps

Transportation maps between probability measures are critical objects in numerous areas of mathematics and applications such as PDE, fluid mechanics, geometry, machine learning, computer science, and economics. Given a pair of source and target measures, one searches for a map that has suitable properties and transports the source measure to the target one. Here, we study maps that possess the \textit{no-collision} property; that is, particles simultaneously traveling from sources to targets in a unit time with uniform velocities do not collide. These maps are particularly relevant for applications in swarm control problems. We characterize these no-collision maps in terms of \textit{half-space preserving} property and establish a direct connection between these maps and \textit{binary-space-partitioning (BSP) tree} structures. Based on this characterization, we provide explicit BSP algorithms, of cost $O(n \log n)$, to construct no-collision maps. Moreover, interpreting these maps as approximations of optimal transportation maps, we find that they succeed in computing nearly optimal maps for $q$-Wasserstein metric ($q=1,2$). In some cases, our maps yield costs that are just a few percent off from being optimal.

math.OC↗

arXiv2019

Nesterov's method with decreasing learning rate leads to accelerated stochastic gradient descent

We present a coupled system of ODEs which, when discretized with a constant time step/learning rate, recovers Nesterov's accelerated gradient descent algorithm. The same ODEs, when discretized with a decreasing learning rate, leads to novel stochastic gradient descent (SGD) algorithms, one in the convex and a second in the strongly convex case. In the strongly convex case, we obtain an algorithm superficially similar to momentum SGD, but with additional terms. In the convex case, we obtain an algorithm with a novel order $k^{3/4}$ learning rate. We prove, extending the Lyapunov function approach from the full gradient case to the stochastic case, that the algorithms converge at the optimal rate for the last iterate of SGD, with rate constants which are better than previously available.

math.OC↗

arXiv2019

The LogBarrier adversarial attack: making effective use of decision boundary information

Adversarial attacks for image classification are small perturbations to images that are designed to cause misclassification by a model. Adversarial attacks formally correspond to an optimization problem: find a minimum norm image perturbation, constrained to cause misclassification. A number of effective attacks have been developed. However, to date, no gradient-based attacks have used best practices from the optimization literature to solve this constrained minimization problem. We design a new untargeted attack, based on these best practices, using the established logarithmic barrier method. On average, our attack distance is similar or better than all state-of-the-art attacks on benchmark datasets (MNIST, CIFAR10, ImageNet-1K). In addition, our method performs significantly better on the most challenging images, those which normally require larger perturbations for misclassification. We employ the LogBarrier attack on several adversarially defended models, and show that it adversarially perturbs all images more efficiently than other attacks: the distance needed to perturb all images is significantly smaller with the LogBarrier attack than with other state-of-the-art attacks.

cs.LG↗

arXiv2019

Calibrated Top-1 Uncertainty estimates for classification by score based models

While the accuracy of modern deep learning models has significantly improved in recent years, the ability of these models to generate uncertainty estimates has not progressed to the same degree. Uncertainty methods are designed to provide an estimate of class probabilities when predicting class assignment. While there are a number of proposed methods for estimating uncertainty, they all suffer from a lack of calibration: predicted probabilities can be off from empirical ones by a few percent or more. By restricting the scope of our predictions to only the probability of Top-1 error, we can decrease the calibration error of existing methods to less than one percent. As a result, the scores of the methods also improve significantly over benchmarks.

stat.ML↗

arXiv2019

Stochastic Gradient Descent with Polyak's Learning Rate

Stochastic gradient descent (SGD) for strongly convex functions converges at the rate $\bO(1/k)$. However, achieving good results in practice requires tuning the parameters (for example the learning rate) of the algorithm. In this paper we propose a generalization of the Polyak step size, used for subgradient methods, to Stochastic gradient descent. We prove a non-asymptotic convergence at the rate $\bO(1/k)$ with a rate constant which can be better than the corresponding rate constant for optimally scheduled SGD. We demonstrate that the method is effective in practice, and on convex optimization problems and on training deep neural networks, and compare to the theoretical rate.

math.OC↗

arXiv2017

Approximate homogenization of convex nonlinear elliptic PDEs

We approximate the homogenization of fully nonlinear, convex, uniformly elliptic Partial Differential Equations in the periodic setting, using a variational formula for the optimal invariant measure, which may be derived via Legendre-Fenchel duality. The variational formula expresses $\bar H$ as an average of the operator against the optimal invariant measure, generalizing the linear case. Several nontrivial analytic formulas for $\bar H$ are obtained. These formulas are compared to numerical simulations, using both PDE and variational methods. We also perform a numerical study of convergence rates for homogenization in the periodic and random setting and compare these to theoretical results.

math.AP↗

arXiv2017

Approximate homogenization of fully nonlinear elliptic PDEs: estimates and numerical results for Pucci type equations

We are interested in the shape of the homogenized operator $\overline F(Q)$ for PDEs which have the structure of a nonlinear Pucci operator. A typical operator is $H^{a_1,a_2}(Q,x) = a_1(x) \lambda_{\min}(Q) + a_2(x)\lambda_{\max}(Q)$. Linearization of the operator leads to a non-divergence form homogenization problem, which can be solved by averaging against the invariant measure. We estimate the error obtained by linearization based on semi-concavity estimates on the nonlinear operator. These estimates show that away from high curvature regions, the linearization can be accurate. Numerical results show that for many values of $Q$, the linearization is highly accurate, and that even near corners, the error can be small (a few percent) even for relatively wide ranges of the coefficients.

math.AP↗

arXiv2017

Approximate Convex Hulls: sketching the convex hull using curvature

Convex hulls are fundamental objects in computational geometry. In moderate dimensions or for large numbers of vertices, computing the convex hull can be impractical due to the computational complexity of convex hull algorithms. In this article we approximate the convex hull in using a scalable algorithm which finds high curvature vertices with high probability. The algorithm is particularly effective for approximating convex hulls which have a relatively small number of extreme points.

cs.CG↗

arXiv2016

A partial differential equation for the strictly quasiconvex envelope

In a series of papers Barron, Goebel, and Jensen studied Partial Differential Equations (PDE)s for quasiconvex (QC) functions \cite{barron2012functions, barron2012quasiconvex,barron2013quasiconvex,barron2013uniqueness}. To overcome the lack of uniqueness for the QC PDE, they introduced a regularization: a PDE for $\e$-robust QC functions, which is well-posed. Building on this work, we introduce a stronger regularization which is amenable to numerical approximation. We build convergent finite difference approximations, comparing the QC envelope and the two regularization. Solutions of this PDE are strictly convex, and smoother than the robust-QC functions.

math.AP↗

arXiv2016

Computing the quasiconvex envelope using a nonlocal line solver

Recently in a series of articles, Barron, Goebel, and Jensen \cite{barron2012functions} \cite{barron2012quasiconvex} \cite{barron2013quasiconvex} \cite{barron2013uniqueness} have studied second order degenerate elliptic PDE and first order nonlocal PDEs for the quasiconvex envelope. Quasiconvex functions are functions whose level sets are convex. The PDE is difficult to solve. In this article we present an algorithm for computing the quasiconvex envelope (QCE) of a given function. The QCE operator is a level set operator, so this algorithm gives a method to compute convex hull of sets represented by a level set functions. We present a nonlocal line solver for the quasiconvex envelope (QCE), based on solving the one dimensional problem on lines. We find an explicit formula for the QCE of a function defined on a line.

math.NA↗

arXiv2016

Numerical methods for motion of level sets by affine curvature

We study numerical methods for the nonlinear partial differential equation that governs the motion of level sets by affine curvature. We show that standard finite difference schemes are nonlinearly unstable. We build convergent finite difference schemes, using the theory of viscosity solutions. We demonstrate that our approximate solutions capture the affine invariance and morphological properties of the evolution. Numerical experiments demonstrate the accuracy and stability of the discretization.

math.NA↗

arXiv2016

Anomaly detection and classification for streaming data using PDEs

Nondominated sorting, also called Pareto Depth Analysis (PDA), is widely used in multi-objective optimization and has recently found important applications in multi-criteria anomaly detection. Recently, a partial differential equation (PDE) continuum limit was discovered for nondominated sorting leading to a very fast approximate sorting algorithm called PDE-based ranking. We propose in this paper a fast real-time streaming version of the PDA algorithm for anomaly detection that exploits the computational advantages of PDE continuum limits. Furthermore, we derive new PDE continuum limits for sorting points within their nondominated layers and show how the new PDEs can be used to classify anomalies based on which criterion was more significantly violated. We also prove statistical convergence rates for PDE-based ranking, and present the results of numerical experiments with both synthetic and real data.

cs.LG↗

arXiv2016

A partial differential equation for the rank one convex envelope

In this article we introduce a Partial Differential Equation (PDE) for the rank one convex envelope. Rank one convex envelopes arise in non-convex vector valued variational problems \cite{BallElasticity, kohn1986optimal1, BallJames87, chipot1988equilibrium}. More generally, we study a PDE for directional convex envelopes, which includes the usual convex envelope \cite{ObermanConvexEnvelope} and the rank one convex envelope as special cases. Existence and uniqueness of viscosity solutions to the PDE is established. Wide stencil elliptic finite difference schemes are built. Convergence of finite difference solutions to the viscosity solution of the PDE is proven. Numerical examples of rank one and other directional convex envelopes are presented. Additionally, laminates are computed from the rank one convex envelope.

math.AP↗

arXiv2015

An efficient linear programming method for Optimal Transportation

An efficient method for computing solutions to the Optimal Transportation (OT) problem with a wide class of cost functions is presented. The standard linear programming (LP) discretization of the continuous problem becomes intractible for moderate grid sizes. A grid refinement method results in a linear cost algorithm. Weak convergence of solutions is stablished. Barycentric projection of transference plans is used to improve the accuracy of solutions. The method is applied to more general problems, including partial optimal transportation, and barycenter problems. Computational examples validate the accuracy and efficiency of the method. Optimal maps between nonconvex domains, partial OT free boundaries, and high accuracy barycenters are presented.

math.NA↗

arXiv2015

Numerical Methods for the 2-Hessian Elliptic Partial Differential Equation

The elliptic 2-Hessian equation is a fully nonlinear partial differential equation (PDE) that is related to intrinsic curvature for three dimensional manifolds. We introduce two numerical methods for this PDE: the first is provably convergent to the viscosity solution, and the second is more accurate, and convergent in practice but lacks a proof. The PDE is elliptic on a restricted set of functions: a convexity type constraint is needed for the ellipticity of the PDE operator. Solutions with both discretizations are obtained using Newton's method. Computational results are presented on a number of exact solutions which range in regularity from smooth to nondifferentiable and in shape from convex to non convex.

math.NA↗