Papers for
statistical modelers
Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.
Bregman geometry reveals new views on exponential family distributions
Bregman geometry in regular exponential families: dual representations and relative smoothness
Abstract: Regular exponential-family distributions are ubiquitous in statistics due to their appealing and convenient properties. Notably, they can be expressed in terms of Bregman divergences, revealing fundamental connections between statistics, convex analysis, and information geometry. However, this geometric structure has not been systematically exploited to characterize these distributions with respect to their expectation parametrization. Building on a projective characterization, we establish a general construction of equivalent dual Bregman representations, together with normalization and uniqueness results. These representations express log-densities as ordinary Bregman divergences in dual coordinate systems, providing geometric characterizations of the parameter space. So-called $β$- and $α$-scaled dual representations emerge as particular instances of this framework. We further derive dual Bregman representations of conjugate priors and establish that they preserve conjugacy. Applications to classical distributions illustrate the geometric structures underlying seemingly different statistical models. Moreover, we show that these representations provide natural reference geometries for establishing relative smoothness of the primal Bregman representation of the distribution, with explicit bounds that are sharp in the scalar and separable settings. An illustrative application to the Poisson distribution sheds new light on several works in the literature, providing complementary geometric justifications for reference geometries while deriving existing algorithms.
Clustering variables in causal graphs improves analysis and clarity
Efficient and Practical Clustering of Variables in Causal Graphs
Abstract: Clustering variables in a graph may improve visual clarity and simplify causal inference. Harnessing the full potential of clustering requires an automated method to detect useful clusters in a graph. We introduce a novel class of clusters called conduits, and present a sound, complete and efficient algorithm for detecting all conduits in the graph. In addition to algorithmic detectability, conduits appear naturally in causal graphs stemming from real-world applications, enabling substantial size reduction. Furthermore, the structure of conduits enables the application of known identifiability results regarding clustering. We demonstrate these properties in a simulation study with real-world graphs.
Learning mixtures of exponential family models with structured labels
Learning Mixtures of Exponential Family
Abstract: Motivated by causal inference and physical computing systems, we consider the problem of learning mixture of two bounded-support exponential-family distributions with a shared interaction parameter matrix and mixture component-specific linear parameter vectors. The unknown sequence of component labels belongs to a known structured class. We propose an estimator based on a convex loss that avoids the log-partition function; for each fixed label assignment, estimation is a convex optimization problem. We show that the conditional population loss uniquely identifies the interaction matrix and the linear parameter attached to each sample. For a global empirical minimizer, we establish finite-sample bounds on the interaction error in Frobenius norm and the average squared linear-parameter error. The bounds depend on the logarithm of the number of admissible label assignments, and, for fixed dimension and model bounds, imply consistency when this complexity grows sublinearly with sample size. Additional occupancy and separation conditions yield component and label recovery.
Improving Markov chain sampling by adding new transitions
Beyond Reweighting: Certified Improvements from New Transitions in Discrete MCMC
Abstract: Optimizing transition probabilities on a fixed grap may leave substantial limitations on the efficiency of reversible Markov chain Monte Carlo (MCMC). We develop a framework for transition-support design that considers both allowed transitions and their stationary flows. For a fixed target distribution, observable, and node-flow budget, we derive a necessary and sufficient dual condition for support augmentation to strictly reduce the optimal value of the asymptotic variance, together with computable primal--dual certificates that establish improvements over every admissible original-support reweighting. We also construct a family in which adding a single edge reduces integrated autocorrelation time by an $Ω(n)$ factor relative to optimal original-support weighting, demonstrating that structural limitations can impose arbitrarily large relative losses in estimation efficiency. To realize these gains, we investigate energy-informed edge selection and constrained stationary-flow optimization. Experiments on molecular and Ising state graphs distinguish the potential value of new transitions from the performance of kernels constructed on them: on identical augmented Ising supports, full weight optimization certifies strict improvements in 24/24 conditions, whereas degree-based Metropolis--Hastings weighting certifies none. Our results provide a principled approach to characterizing, certifying, and exploiting transition-support improvements in explicitly represented finite state spaces.
Entropy and divergence concepts extended to non-invariant probabilities
Thermodynamic Formalism for Non-Invariant Probabilities: Entropy, Divergence, Channels
Abstract: Let $\mathcal{M}$ and $\mathcal{M}_σ$ denote, respectively, the sets of Borel and $σ$-invariant probabilities on the uncountable set $Ω=\{1,2,...,d\}^\mathbb{N}$. Given a Hölder Jacobian $J$, the dual Ruelle operator $μ\in \mathcal{M} \mapsto \mathcal{L}_{J}^* (μ)$ will play the role of a dynamical channel; generalizes the action of a stochastic matrix on probability vectors. Since $\mathcal{L}_{J}^*$ generally breaks $σ$-invariance, we introduce the \emph{ni-entropy} $h_{ni}(μ)$ for non-invariant measures, employing the concept of the expected value function on cylinders of size one. $h_{ni}(μ)$ differs from Shannon entropy but extends the Kolmogorov-Sinai entropy from $\mathcal{M}_σ$ to $\mathcal{M}$; $h_{ni}$ is concave and u.s.c. This leads us to the concept of ni-pressure $\mathcal{P}(A)$ of a potential $A:Ω\to \mathbb{R}$. $μ\in \mathcal{M}$ is not a density with respect to a dominating measure. We define an {\it ni-Kullback-Leibler divergence} $D_{ni}(ρ|| ν)$ not assuming absolute continuity. We investigate concepts of Information Theory such as Minimal Output Entropy, Donsker-Varadhan variational principle, a variational formula for $h_{ni}$, bounds for divergence, information projections, and their connections to the ni-pressure $\mathcal{P}$.
Clique constraints improve exact causal learning efficiency in some cases
Clique Constraints for Exact Causal Structure Learning
Abstract: Exact score-based causal structure learning supports explicit structural constraints and optimisation certificates, but its branch-and-cut search can be computationally demanding. We augment exact causal structure learning methods with an upper bound on the maximum clique size of the learned undirected skeleton, enforced at integer candidates through lazy cycle and clique constraints. On static linear-Gaussian Erdős-Rényi and scale-free DAGs, we evaluate callback order, alternative clique-cut policies, incumbent repair, fractional separation, and Big-M bounds. We report causal-recovery metrics jointly with optimality gaps, primal and dual bounds, node throughput, MIPSOL events, and callback cost. Two variants receive particular attention: B3 selects the most violated minimal (k+1)-clique inequality at each callback, whereas C3 repairs an infeasible integer candidate by deleting violating edges and refitting affected regression columns using QR-based least squares. The results do not indicate that clique constraints uniformly accelerate optimisation. However, their benefit depends on graph density, sample size, and cut-management policy, while large collections of correlated constraints can reduce node throughput. Tighter valid Big-M bounds reduced the fixed-budget optimality gap in most tested settings, without necessarily changing SHD or F1 at a fixed time limit. These findings characterize the computational trade-offs of clique-constrained exact causal learning, and identify selective separation and incumbent repair as promising directions.
Polynomial density neurons clarify exact outputs for missing inputs
Evidence or Marginal? What Moment-Vector Propagation Computes in Polynomial Density Neurons
Abstract: When a neuron's input is lost, or known through a belief $q$, what should it output? For a neuron storing a joint polynomial density $N$ with input density $D$, two formulas compete. $p_\mathrm{CD}=E_q[N]/E_q[D]$ costs one matrix-vector product; Hierarchical Correlation Reconstruction introduces it as a constant-denominator approximation to $p_\mathrm{EU}=E_q[N/D]$. For a density model both are exact answers to different questions: $p_\mathrm{CD}$ is Bayesian updating with the belief as virtual evidence; $p_\mathrm{EU}$ installs the belief as a new marginal (Jeffrey's rule). With a missing input $p_\mathrm{CD}$ returns the model's conditional; $p_\mathrm{EU}$ generally does not. This post hoc reading is proved, and illustrated outside the pair family up to $k=8$ inputs; it has no pre-registered test, only one exploratory real-data pilot. Post hoc, closed forms fall within a factor 2 of all 112 measured gaps, and negative density begins near input correlation $1/k$. In a pre-registered synthetic study (up to 32 inputs), 5 of 15 hypotheses are supported, 2 inconclusive and 8 not supported, including all three on transfer beyond equicorrelated Gaussians. Post hoc, where $p_\mathrm{EU}$ is defined, truncation usually outweighs the gap for Beta beliefs in pair neurons; triple neurons under flat and data beliefs are exceptions.
Improve sampling in complex high dimensional models with geometry aware monte carlo
Geometry-Aware Score-Repellent Monte Carlo
Abstract: The recently proposed Score-Repellent Monte Carlo (SRMC) method improves MCMC sampling by tilting the target density $π$ in $\mathbb{R}^d$ with the factor $\exp(-α\,θ_n^\top s(x))$, where the score function $s(x) = \nabla_x \log π(x)$, and $θ_n$ is the score average from past samples. However, a single scalar repellence strength $α\ge0$ cannot handle both the steep and the flat directions in anisotropic energy landscapes at once. We introduce \emph{Geometry-Aware SRMC} (GA-SRMC), which generalizes the Euclidean alignment to $θ^\top M s(x)$ for any symmetric positive-definite matrix $M$. A stochastic-approximation central limit theorem under a trace constraint on $M$ identifies the trace-normalized inverse Fisher $M_{\rm opt}\propto S^{-1}$ as the unique minimizer of the worst-direction covariance bound, elevating the Fisher information $S=\text{Cov}_π(s,s)$ from a proposal-side preconditioner (e.g., FisherMALA, natural gradient) to the bound-optimal repulsion-side matrix, with a single online estimate $\widehat S_n$ serving both roles at no additional computational cost. The full-Fisher online metric $M_n$ converges almost surely to $M_{\rm opt}$ and satisfies a CLT, providing a theoretical guarantee for the online implementation of the GA-SRMC framework. On an $8$-target simulation spanning the condition number $κ\in[1.3,100]$ and dimension $d\in[2,785]$, GA-SRMC dominates SRMC in ESS by $\boldsymbol{4.3\times}$ on anisotropic Gaussian ($κ=100$) and $\boldsymbol{6.8\times}$ on MNIST under the MALA baseline, by $\boldsymbol{16\text{--}32\%}$ on three well-conditioned logistic-regression posteriors, and by up to $\boldsymbol{+48\%}$ under FisherMALA; gains peak on the most anisotropic targets, and posterior-mean MSE is lower than with SRMC in all $16$ full-Fisher implementations.
Query limits exposed for advanced sampling from complex distributions
Towards the Query Complexity of Sampling under Strong Log-Concavity and Log-Sobolev Inequalities
Abstract: We study the query complexity of sampling from smooth distributions on $\mathbb R^d$ using first-order oracles. For $d$-dimensional strongly log-concave targets with condition number $κ=2$, we prove that any randomized sampler requires $\exp({Ω(\log^{1/3} d)})$ queries even from a constant warm start. This rules out polylogarithmic dimension dependence and strictly separates strongly log-concave sampling from Gaussian sampling. For targets with $L$-smooth potentials satisfying an $α$-log-Sobolev inequality with $κ= L/α$, we prove an $Ω(κ)$ query lower bound for $d\gtrsim\log^2κ$. This shows that log-Sobolev sampling is strictly harder than strongly log-concave sampling. Our proofs construct instance-dependent tests that target samples typically pass but outputs of algorithms making few queries typically fail. For strongly log-concave targets, we embed a sequence of hidden Gaussian into high-degree tensor polynomial with smooth gates, forcing any algorithm to uncover the layers sequentially. The log-Sobolev construction hides a constant fraction of the target mass in angular caps while controlling the log-Sobolev constant via delicate estimates.
Estimation risks clarified when nuisance parameters vanish
Exact Global Minimax Risk near Nonidentification
Abstract: We study squared-error estimation when a nuisance parameter is unidentified on a submodel but affects a target that vanishes there. For models with higher-order profile expansions, we give sufficient conditions under which the largest critical-path minimax risk is the exact asymptotic minimax risk over the full parameter space. An efficient observable feature, negligible calibration error and stable recovery yield one estimator attaining the leading constant uniformly, with the profile coordinates unknown. The estimator uses calibration estimates, a local feature mean and an empirical recovery feature. The constant is expressed through a bounded-normal decision problem. We verify the conditions in two models. In centered Gaussian logistic errors-in-variables regression with one surrogate and two unknown variances, we determine the fifth-order global coefficient risk and its exact constant. A sparse extension separates variance-decomposition uncertainty from direction estimation and specifies when the scalar constant is preserved. In a degree-balanced replicated-network model, we determine the sixth-order global edge-contrast risk when block proportions are unknown. Triangles identify a parameter combination; a bowtie supplies the additional information needed for the contrast.
Stationary infinite memory chains yield adaptive empirical confidence bounds
Empirical concentration for chains with infinite memory
Abstract: Concentration inequalities for time series usually involve unknown quantities measuring dependence. This prevents their direct use as confidence bounds, or as criteria for learning algorithms. Recent works proved empirical concentration inequalities for stationary Markov chains, where the unknown quantities are estimated. We extend this approach beyond the Markov case, by considering stationary chains with infinite memory. We first establish a bounded-differences inequality on bounded Polish metric spaces under a summable Wasserstein contraction condition. We then estimate its dependence coefficients from a single trajectory, using either transition counts when the state space is finite, or derivatives of kernel conditional-density estimates for a continuous state space. We give general error bounds and illustrate their dependence on the memory tail. Finally, a fixed multiplicative safety factor turns these estimators into empirical concentration bounds. While some regularity assumptions are required to make the bounds empirical, our procedure remains fully adaptive in the sense that it does not require the knowledge of the constants involved in these assumptions. The empirical confidence bounds derived from our results are asymptotically as tight as the theoretical ones.
NPMLE achieves optimal rates for estimating Gaussian mixtures
Minimax Optimality of the NPMLE for Gaussian Mixtures in Hellinger Distance
Abstract: We analyze the problem of estimating univariate Gaussian mixture densities using the nonparametric maximum likelihood estimator (NPMLE) based on $n$ independent and identically distributed samples. We show that the unconstrained NPMLE achieves the minimax optimal rate $\log(n)/n$ in squared Hellinger distance, when the true mixing measure is sub-Gaussian. In addition, we show that when the true mixing measure has compact support, the NMPLE constrained to a compact interval also achieves the minimax optimal rate, which is slightly faster and of order $\log(n)/(n \log \log n)$. Our proof consists of sharply characterizing the local Hellinger bracketing entropy of the class of finite Gaussian mixtures. One of our key technical innovations is to sharply bound the $L^2$ bracketing entropy of balls of bounded-degree polynomials, which may be of independent interest. As an application, we establish the regret-optimality of $g$-modeling procedures based on the NPMLE in the Gaussian Empirical Bayes problem with bounded priors.
Binomial distribution has highest uncertainty among certain dependent sums
The binomial distribution maximises entropy among sums of negatively related indicators with fixed mean and support
Abstract: We show that the binomial distribution supported on $\{0,1,\ldots,n\}$ and with mean $np$ for some $0<p<1$ maximises entropy within the class of sums of negatively related indicator random variables with this mean and support. Our results generalise the main result of Yu (2008), who showed that this binomial distribution maximises entropy within the smaller class of ultra-log-concave random variables of order $n$. Our proof makes use of a Markov chain construction of Yu (2008), together with a stochastic ordering argument.
Sample size requirements grow rapidly for weak latent-variable signals
Sample complexity of weak-signal latent-variable models
Abstract: We study parametric estimation in latent-variable models in the weak-signal regime, where a small signal-strength parameter $β$ controls the dependence of the observations on the unknown parameter. At $β=0$, this dependence disappears, so all parameter values induce the same observed distribution, which we call the null law. We develop a unified framework for analyzing sample complexity in the weak-signal regime by expanding the likelihood ratio around this shared null law in powers of $β$. We prove that if the unknown parameter is first globally identifiable at order $r$ in this expansion, then estimation to fixed accuracy requires at least on the order of $β^{-2r}$ samples. We further identify a likelihood-separation stability condition that is necessary for any estimator to attain this rate and, under mild regularity conditions, sufficient for both maximum likelihood and method-of-moments estimators to attain it. We also study the regime in which the target accuracy vanishes together with the signal strength, and show that global localization and local refinement may be governed by different likelihood orders, leading to distinct statistical scales. Finally, we apply our framework to broad classes of algebraically structured latent-variable models, including additive-noise channels and exponential-family models.
ReLU networks approximate smooth functions with optimal error rates
Optimal approximation of smooth functions by ReLU neural networks
Abstract: We study uniform approximation of the Hölder ball $\mathcal H_1^s([0,1]^D)$ by fully connected ReLU networks. Except in the one-dimensional Lipschitz case, existing constructions remain separated from the information-theoretic efficiency limit. This article closes the gap for every $s>0$ and $D\ge1$: networks with width $O(N)$, depth $O(L)$, and parameter magnitude $O(1)$ attain uniform approximation error of order $(N^2L^2\log N)^{-s/D}$. No relation between width and depth is required. The construction uses new encoding and decoding mechanisms that exploit higher-order finite-difference bounds for local Taylor coefficients of the target function. As a statistical consequence, empirical risk minimizers over the corresponding clipped network classes attain the minimax regression rate $n^{-2s/(2s+D)}$ without additional logarithmic factors. The paper also gives a detailed exposition of the key approximation techniques, covering both established and new mechanisms.
Wide neural networks converge functionally at rate inverse square root of width
Lifting Finite-Sample Gaussian Approximation of Wide Neural Networks to Functional Convergence
Abstract: Randomly initialised neural networks of width $m$ converge to Gaussian processes as $m \to \infty$, with two types of quantitative bounds characterising this convergence. Finite-sample bounds compare the network with its Gaussian limit at a fixed finite sample set, on an arbitrary domain, in the quadratic Wasserstein distance $\mathcal W_2$, achieving a convergence rate of $m^{-1/2}$ with explicit constants. However, these constants grow with the number of samples, so the bound becomes uninformative as the sample set is refined. Functional bounds remove that dependence by comparing the two as random functions, at the price of spherical input, metrics weaker than $\mathcal W_2$, or slower rates. In this paper, we therefore lift a finite-sample bound to a functional one. The lifting refines the sample set until it is dense in the domain, so the bound must first be made independent of the number of samples. We achieve this by measuring the error per sample, which makes the constant uniform. For a polynomial activation and any compact input set, this gives a rate of $m^{-1/2}$ in the supremum norm, with an explicit constant.
DisCo algebra reveals limits of summarizing joint probabilities
DisCo Algebra: Why Disintegration and Combination Are Not Inverses
Abstract: We present DisCo algebra---a compositional framework for joint probability built from a symmetric cooperad (disintegration) and a symmetric operad (combination) on the bare family $\mathcal{P}_n=\mathrm{Prob}([0,1]^n)$. No probability integral transform, no external summary functions, no enrichment of objects with conditional kernels. The disintegration side uses block-wise joint CDFs as intrinsic, measure-determined 1D summaries, yielding a lossy compression. The combination side reads the same summaries and reconstructs via regular conditional measures, yielding a lossless embedding. We prove that combination followed by disintegration is the identity on admissible inputs, but disintegration followed by combination is not---the information loss is exactly the conditional KL divergence between the true conditional distribution and the conditionally independent product. This asymmetry reveals a hard limit on how much cross-block dependence can be captured by scalar block summaries.
Kernel method tests how well conditional models match real data
Conditional Kernel Stein Discrepancy
Abstract: Kernel Stein discrepancies (KSDs) provide a versatile tool for comparing distributions. One of their main applications is in quantifying the goodness-of-fit (GoF) between a data-generating distribution and a prescribed target distribution. In this work, we study the related problem of conditional GoF quantification: given only a (possibly non-normalized) conditional target model, without information on the distribution of its covariates, and samples from a joint distribution, the goal is to assess how well the conditional distribution of the samples matches the target. To tackle this setting, we present a framework that allows lifting unconditional KSDs to the conditional setting through an operator-valued kernel on the covariate space, going beyond the known Euclidean case. We establish that our suggested statistic vanishes if and only if the conditional model and the true conditional distribution agree for almost all covariates and deploy it to test conditional GoF on smooth manifolds and on discrete spaces. Our experiments on level, power, and runtime demonstrate the viability of testing on these domains using the proposed statistic.
Symbolic expression learning improves understanding of complex unnormalized distributions
Symbolic Density Estimators for Unnormalized Distributions
Abstract: Estimating the symbolic or analytical form of probability density functions (PDFs) from observed samples is a fundamental challenge in statistical and computational modelling. This process is critical for deriving interpretable and generalizable relationships characterizing the underlying phenomenon. Traditionally, this estimation depends strongly on domain expertise and prior field-specific knowledge, with experts selecting appropriate functional forms or parametric families based on empirical evidence and theoretical understanding. The coefficients of these forms are then typically determined through parameter estimation. In this paper, we develop a framework for estimating symbolic expressions of unnormalized distributions from observed samples using domain-specific prior knowledge, such as the range of interactions and a predefined set of primitive functions. We integrate deep generative models with symbolic regression (SR), incorporating inductive biases, such as factorizing large distributions, to keep the problem tractable. The deep generative models we examine include likelihood-based models, viz., flow models, and score-based models. Experiments show the effectiveness of the proposed framework for estimating density functions for multivariate toy distributions as well as lattices from computational physics, namely, XY model and $φ^4$ theory. When applied to the renormalization problem in $φ^4$ theory, the proposed framework estimates compact symbolic approximations of the hamiltonian function at different scales directly from samples, yielding expressions that may be challenging to derive using traditional perturbative or analytic approaches in nonperturbative settings.
Gaussian process designs reveal what model assumptions can be tested
What Can a Gaussian Process Design Test
Abstract: A Gaussian process (GP) model can agree with the data for two reasons: its assumptions are right, or the chosen inputs could never have shown that they are wrong. The distinction can be checked from the design before any responses are observed. Every model implies relations that its noiseless responses must satisfy at the chosen inputs, such as the middle value lies on the line through its two neighbours. For GPs built from finitely many features, these relations are exactly the null space of the kernel matrix. Gale duality gives them a geometric interpretation, in which each observation has a vector and the smallest groups of observations that can expose an error are the circuits. For other kernels the relations become soft: response patterns may be improbable under the prior rather than algebraically impossible. A standard test then combines two kinds of evidence. Structural evidence comes from a violated relation and grows without limit as the noise falls. Prior-based evidence only says that a departure is improbable under the prior. With all inputs at the two ends of an interval, for example, a GP can reject a straight line against a large curvature, but only because the implied intercept is improbable, never because curvature was seen. In simulations the predicted power matched the observed rejection rates. Choosing the next input by predicted power raised the power against a localised discrepancy from 0.48 to 0.72, against 0.51 when choosing by predictive variance, and a grid in two dimensions contained exact tests of additivity that a Latin hypercube lacked. The test itself is classical. The contribution is the prospective reading of that test: before observing the responses, the design already determines what kind of contradiction it can produce.
Adaptive method improves causal discovery with few data points
AdaPS-LiNGAM: Adaptive Predecessor Selection for Linear Non-Gaussian Acyclic Models under Small-Sample Settings
Abstract: Causal discovery becomes particularly challenging when the available sample size is small relative to the number of variables. This challenge also arises in the linear non-Gaussian acyclic model (LiNGAM), an identifiable framework for causal discovery from observational data. DirectLiNGAM estimates a causal order, which arranges variables so that causes precede their effects, by sequentially identifying an exogenous variable and removing its linear effect from the remaining variables. We establish a structural limitation of this procedure: when the number of variables exceeds the sample size, repeated residualization necessarily becomes degenerate before the full causal order can be determined. Our analysis further reveals that each residual can be reconstructed using only a graph-determined subset of variables already placed earlier in the causal order, termed the active boundary. This result motivates AdaPS-LiNGAM (Adaptive Predecessor Selection LiNGAM), which reconstructs each residual directly from the original observations using an adaptively chosen sparse subset of those earlier variables. The same subset-selection principle is also applied to the final pruning step for edge estimation. Experiments on synthetic data demonstrate that AdaPS-LiNGAM provides accurate causal-structure recovery in sample-limited settings and degrades more gradually as the sample size decreases.
Mean field fluctuations reveal uncertainty in neural network training outputs
Fluctuations of Nonlinear Observables in Mean Field Neural Network Training
Abstract: Mean field limits describe the training dynamics of wide neural networks through the evolution of the empirical distribution of their parameters. Although functional central limit theorems characterize the asymptotic fluctuations of this distribution, quantities of practical interest are typically nonlinear observables of the parameter distribution rather than the distribution itself. In this work, we show how these mean field fluctuations propagate to finite dimensional nonlinear observables for shallow neural networks trained by stochastic gradient descent. Working in the weighted Sobolev space in which the limiting fluctuation process is constructed, we apply a functional Delta method under ordinary Fr{é}chet differentiability, without requiring Lions derivatives with respect to the measure variable. We obtain a central limit theorem for the observables and, under a suitable representation of their differentials, an explicit covariance formula inherited from the underlying mean field fluctuation theory. We also study whether prescribed quantities of interest can be recovered from the selected observations. Under a constant rank assumption, we prove that a quantity of interest factors locally through the observation functional if and only if, throughout a neighborhood, the kernel of the differential of the observation is contained in that of the quantity of interest. Thus, a differential condition expressed directly in the ambient Sobolev space yields an exact nonlinear local factorization. These results provide a framework both for quantifying finite-width uncertainty on observable, statistically or physically meaningful quantities and for assessing whether the chosen observations contain the information required to identify them.
Reflected anchored Langevin algorithm samples nonsmooth constrained distributions
Reflected Anchored Langevin Algorithms
Abstract: First order Langevin algorithms for constrained sampling in machine learning, such as projected Langevin Monte Carlo which are based on discretizations of reflected Langevin dynamics, require differentiable log densities that limits their applicability. This paper introduces reflected anchored Langevin dynamics (RALD), a reflected diffusion that converges to non-differentiable targets on constrained domains. The method uses a smooth anchored reference potential and multiplies the drift and noise covariance of its reflected Langevin dynamics by the same state dependent scaling factor. Its Euler-Maruyama discretization with projection gives reflected anchored Langevin Monte Carlo (RALMC) algorithm. We prove explicit convergence bounds and iteration complexity for RALMC in the 2-Wasserstein distance to the target distribution. Numerical experiments are provided to illustrate the theoretical predictions and the empirical performance of the method.
Likelihood approximation using chained normalizing flows for simulation inference
A perspective note on likelihood approximation and inference for complex simulation models using a chain of aggregated normalizing flows
Abstract: We present a new perspective on the problem of likelihood approximation within the framework of simulation-based inference that promotes scalable and controllable simulation routines for large-scale data analysis, allows efficient parameter space exploration or smooth interpolation in high-dimensions and, thus, supports valid statistical treatments of hypothesis testings as well as uncertainty quantification. In particular, we consider a chain of $n$-aggregated normalizing flows for likelihood approximation scheme, where a set of upfront replicated observation datasets from the forward complex simulation model pass through the first set of bijective transformations, and then subsequently pass to the other sets of bijective transformations. Here, we assume that, for any $k \in \{1,\,2, \ldots, n\}$, the parameters corresponding to the first $k$ sets of bijective transformations are estimated sequentially, in some sense of optimality, for constructing flexible probability distributions, regardless of the remaining $(n-k)$ sets of bijective transformations. Moreover, our objects of interest are to highlight two complementary mathematical arguments that leverage an informatics-theoretic formalization, based-on empirical likelihood estimators under moment restrictions, and a sequential decision-making paradigm, with mixing distributions, for updating and aggregating the estimated parameters of the overall normalizing flows. As a by-product, the framework provides a reliable surrogate model, conditioned on the model parameters defining the forward computational simulation, that allows samples generation, with statistical powers, and facilitates computationally tractable scheme in the Bayesian paradigm for inference, hypothesis testings and uncertainty quantification.
Guidance estimation improves steering of diffusion model sampling
WEIRDO: WEak resIdual Regularized DOob's h-transform diffusion alignment
Abstract: We study the problem of estimating the guidance that steers the distribution learned by a diffusion generative model toward a tilted target $q_0 \propto w\,p_0$ at inference time. Relying on the stochastic optimal control approach, we observe that the exact drift correction is the gradient of the logarithm of Doob's $h$-function, and we study the problem of estimating it from a sample. In the present paper, we assume that the score of the pretrained model is available, that the tilting weight is bounded and positive, and that the reference distribution has a bounded support, no smoothness of the weight is required. Introducing a penalized least-squares risk in which the penalty is the residual of the space-time harmonicity equation satisfied by the $h$-function, measured in a dual Sobolev norm, we derive high-probability bounds on the squared error of the resulting guidance estimate. Since the penalty vanishes at the target, the estimator is free of regularization bias, and in favourable scenarios its rate of convergence is faster than the minimax rate of estimating first-order derivatives of a smooth regression function. Assuming that $w$ is bounded and positive with $\mathbb{E}_{p_0}[w^{-\mathrm{s}}] < \infty$ for some $\mathrm{s} \in (0,\infty]$, and that the reference data are compactly supported, we prove that the guidance is estimable in squared $L^2$ at rate $\varepsilon_n^{\mathrm{s}/(\mathrm{s}+4)}$, where $\varepsilon_n = n^{-2(β-1)/(2(β-1)+d)}.$ We also transfer the obtained bounds to the total variation distance between the marginals of the estimated and the exactly guided samplers, and illustrate the performance of the suggested approach with numerical experiments.
Information splits explain probability deviation events geometry
An information identity reveals the geometry of deviation events
Abstract: In a deviation event, the empirical measure of $n$ independent draws lands in a set of distributions. The exponent of its probability splits, at every $n$, into two nonnegative terms. The first is the rate: $n$ times the relative entropy, from the population, of the distribution of a typical draw under the event. The second is the dependence that conditioning induces among the draws, measured by their total correlation. Their relative sizes depend on the geometry of the set. This paper focuses on this exact split and on the geometry that the interplay of its two terms reveals. The dependence vanishes exactly when the event confines every draw to a single set, and the classical rate of that constraint is exact. On a set invariant under a compact group that fixes the population and leaves no invariant set of intermediate probability, the rate vanishes; uniform convergence over a hypothesis class that such a group permutes is one. On a half-space whose threshold sits a fixed number of standard errors above the mean, the fraction of the exponent that is dependence tends to a function of the event's probability alone. On a set split into pieces, the dependence is the pieces' average plus $n$ times the Jensen-Shannon divergence of their marginals, minus the entropy of their weights. The split extends to a general reference law. When the information projection is such a law, the split gives the exact value of the expectation in the exponential change of measure to the projection.
Inferring dynamic changes from distribution snapshots requires multiple time points
Beyond Gradient Flow: Identifiability and Recovery from Distribution Snapshots
Abstract: Inferring dynamics from snapshots of evolving distributions is fundamentally underdetermined: the Fokker-Planck equation constrains the drift $F$ only through its score-weighted divergence $\nabla\cdot F+F\cdot\nabla\logρ$, leaving a $ρ$-solenoidal gauge invisible to any single-time constraint. Time-indexed transport formulations cannot resolve this ambiguity: every admissible marginal path admits a curl-free explanation, minimum-action reconstruction selects it, and marginal fit alone cannot distinguish dynamically inequivalent explanations. Requiring one autonomous field to explain several marginals instead makes part of the hidden circulation visible as $\nabla\logρ$ changes across marginals. Separating instantaneous Fokker-Planck source constraints from the snapshot experiment, we show that the source constraints identify the field modulo the kernel of a stacked score-weighted divergence operator. For generic Gaussian shape variation, source constraints at $K\ge m$ time points in intrinsic dimension $m$ eliminate every polynomial gauge direction, whereas finitely many density snapshots alone admit aliasing; we give the obstruction explicitly. At a Gaussian anchor, for Sobolev smoothness $s$ and $n$ samples per time point, we derive a conditional lower rate $(nK)^{-2s/(2s+m+1)}$ for the tangent snapshot experiment, with a matching upper rate in a degreewise benchmark. Strong-form fitting is non-orthogonal to score error and cannot be repaired by spectral filtering. Instead, we estimate using smooth test functions while retaining the known diffusion term, and derive a finite-sample bound that separates sampling error from fixed-grid quadrature bias. Planted-circulation experiments confirm the predicted gauge contraction and expose a design tension between cross-slice information and covariance-aware whitening.
New method finds lower-dimensional shape behind complex data sets
Recovering Lower-Dimensional Semialgebraic Support of a Measure from its Moments
Abstract: Recovering probability measures from their moments has numerous applications, esp. in connection with the method of moments in statistics and optimization. In the setting where measure need not be finitely atomic, but its support is known to be compact and semialgebraic with codimension at least one, the problem is still open. We combine moment-matrix kernel information with the Christoffel--Darboux kernel to provide a discrete approximation of the support. To validate the proposed approach, we test our algorithm on analytically computed moments and pseudo-moments arising from polynomial optimization problems without unique global minimizers. This complements well-known recent work on recovery of measures with algebraic support, where the kernel of a moment matrix can reveal polynomials vanishing on the support, and on recovery of sufficiently regular full-dimensional supports, where estimators constructed by thresholding the Christoffel--Darboux kernel are known to converge asymptotically to the support.
Normalizing flows learn complex distributions without example samples
SIMANF: Sample Free Learning of Unnormalized Distributions via Simulated Annealing in Normalizing Flows
Abstract: Efficiently learning and sampling from high dimensional, multimodal unnormalized distributions without target samples remains a challenging problem. Although normalizing flows can generate samples efficiently, training based on the reverse KL divergence using only the unnormalized target density may suffer from mode collapse. We introduce SIMANF, a sample free framework that integrates simulated annealing with normalizing flows. SIMANF progressively transforms the target distribution from a smooth initial form to the original target distribution and trains the flow sequentially across these stages. By transferring the learned representation between stages, the method promotes mode coverage while progressively capturing finer features of the target distribution. Following annealing, a final refinement stage combines the reverse KL divergence with an importance weighted forward KL objective using samples generated by the flow. SIMANF requires no target samples during training and uses only the unnormalized density. We demonstrate its effectiveness on Many-Well distributions and high dimensional Scalar Phi4 lattice field theory distribution.
Formula reveals loss patterns in shallow relu networks with bias
Population loss in shallow ReLU networks: Bias & families of critical points
Abstract: The main result presented is a formula for the population loss in the student-teacher kernel model that is applicable to shallow ReLU networks with bias. This extends previous work of Choo and Saul (2009) and Brutzkus and Globerson (2017). The formula makes essential use of Owen's T-function. The necessary theory of the T-function is given and a high precision coding using MPFR for the T-function, based on an algorithm of Komelj (2023), is available on request. It is shown that various families of spurious minima described in past papers of Arjevani and the author extend to biased networks and that the loss is always strictly decreased when bias is added. The change in landscape geometry caused by adding bias appears to be relatively mild. Only the simplest examples are described in this paper where it is assumed that the number of inputs is equal to the number of neurons (this restriction is for reasons of length). A review of relevant previous results on unbiased networks is included. Aside from Gaussian statistics, the main mathematical tools and ideas come from analytic geometry (analytic and subanalytic sets, the Curve Selection Lemma).
Discrete diffusion model improves simulation of spin systems and sampling
Discrete Diffusion Models via Evolving Variational Autoregressive Networks
Abstract: Conventional score-based diffusion models learn scores without representing normalized densities, whereas tractable normalized models support both sampling and direct likelihood evaluation. A recent tensor-network approach provides such a representation but is largely restricted to low-dimensional lattices. Here we introduce a discrete diffusion model that parameterizes normalized probability distributions using variational autoregressive networks. Explicit Markov jump operators govern the forward noising and reverse denoising dynamics, extending discrete diffusion models with normalized distributions to spin systems on higher-dimensional lattices. We apply this framework to the two- and three-dimensional Ising models across ordered, critical, and disordered regimes, accurately computing thermodynamic quantities including free energy, energy, and magnetization. We further integrate the framework with Monte Carlo sampling, using adaptive diffusion steps to maintain high acceptance rates even at low temperatures while enhancing sample diversity. These results establish a neural-network framework for the discrete diffusion model with normalized probability distributions.
Copula operad framework links dependence and entropy additively
Copula Operad and Copula Entropy
Abstract: We construct a symmetric operad $\mathfrak{C}$ on the class of all multivariate copulas, with operadic composition given by Sklar substitution. This places the hierarchical combination of dependence structures into an algebraic framework. Restricting to absolutely continuous copulas whose densities lie in $L\log L$, we prove that copula entropy is additive under composition, making it an additive character on suitable finite-entropy suboperads. We exhibit explicit closed suboperads (bounded, $L^{p}$, boundary-growth) and show by counterexample that componentwise $L\log L$ does not imply closure under composition.
MaxEnt probability distributions on spheres inform models of cognition
Maximum Entropy Probability Distributions on Spheres with Fixed Mean Busemann Function and Holomorphic-Information-Geometric Model of Cognition
Abstract: In the first half of the paper we revisit the question regarding MaxEnt probability distributions on spheres. We derive families of MaxEnt distributions on spheres in real and complex vector spaces with fixed expected Busemann function (energy). As a particular case, we deduce sub-families on canonical energy levels where inverse temperature equals the dimension of the sphere. In the second part we focus on the information manifold of canonical MaxEnt distributions on the sphere in the complex vector space. This manifold is isomorphic to the Bergman ball. We introduce the reproducing kernel on this manifold and use the RKHS theory to elaborate the model of cognition. In particular, we state the principle of minimal cognitive effort in RKHS and demonstrate its dual relationship with the MaxEnt principle for probability distributions on the boundary sphere.
Compressed subspaces speed up uncertainty analysis in large models
Compressed Active Subspaces for Scalable Bayesian Inference
Abstract: Active subspace methods provide a framework for quantifying predictive uncertainty in high-dimensional models by identifying and performing inference along parameter directions that have the greatest influence on the model output. However, the construction of active subspaces requires storing many full-dimensional model gradients, which becomes prohibitive as model size increases. We address this limitation by proposing Compressed Active Subspaces (CAS), a scalable approach that first maps the model parameters to a compressed space using a structured isometric embedding and then constructs the active subspace within this reduced parameterization. Our approach substantially reduces the memory required for active subspace construction and enables Bayesian inference for large models where standard active subspace methods become impractical. We demonstrate the scalability of CAS on neural networks of increasing size while maintaining predictive performance and robust uncertainty estimates.
Probability distribution evolution explained through least action principle
Physics of Information Geometry - Part I: Principle of Least Action on the Probability Simplex
Abstract: We develop a least-action framework for describing how a probability distribution can evolve from an equilibrium state to a prescribed nonequilibrium state under constrained incremental changes. Taking a Gibbs distribution as the equilibrium reference, the framework gives a direct physical meaning to the geometry of the probability simplex: distance from equilibrium corresponds to nonequilibrium free energy, while changes between successive distributions carry an informational kinetic cost. The Pythagorean structure of relative entropy then provides the central insight of the work. It shows that intermediate distributions chosen via sequential information projections can reduce the kinetic cost of large transitions and establishes an energy-conservation-like relation between the kinetic expenditure along a path and the free energy accumulated in reaching the target distribution. Motivated by this geometry, we construct a greedy least-action path through successive information projections, obtain a closed-form characterization of each projection through the Lambert W function, and establish a finite-step performance guarantee. We further show that state-dependent costs can be incorporated naturally by reshaping the underlying Gibbs reference, providing a thermodynamic interpretation of path penalties as modifications of the effective energy landscape. Together, these results provide a unified view of distributional evolution through least action, information geometry, and nonequilibrium thermodynamics.
Markov chains produce work only with non-Gaussian models in Bayesian inference
Thermodynamic Cyclic Processes with Markov Samplers in Bayesian Inference
Abstract: The concept of Markov chain Monte Carlo (MCMC) cycles, an analogy to cyclic processes in heat engines, is presented in order to examine Bayesian inference problems. In this effort, we develop adaptive ensemble schedulers that allow the tuning of external parameters of a Bayesian canonical ensemble during an MCMC run, realising the MCMC cycles in practice. We run these cycles on different statistical models. As a fundamental insight, we find (both theoretically and in practice) that such systems can produce a non-zero net work output if and only if the considered model is non-Gaussian. As such, they may serve as a measure of non-Gaussianity in Bayesian inference, which we test on an example from supernova cosmology.