Langevin dynamics fails to produce accurate samples even with small score function errors.
problem Robustness of Langevin dynamics to score function errors.
method Analysis of Langevin dynamics and score function errors.
result Langevin dynamics produces a distribution far from the target distribution in TV distance even with small L2 errors in the score function. Study finds simple model-agreement scores perform well in various error estimation scenarios.
problem Evaluating model performance on unseen distributions using disparate scoring functions.
method Rigorously studied popular scoring functions (confidence, local manifold smoothness, model agreement) independently of mechanism choice.
result Simple model-agreement scores outperform confidence- and smoothness-based scores in realistic settings with compromised training data.
New geometric analysis shows L2 score error is flawed for diffusion models.
problem Score matching errors in diffusion models do not fully capture distributional quality.
method Decomposed score errors into gradient and solenoidal components, focusing on gradient's role in Fokker-Planck dynamics.
result Only gradient component affects marginal distributional quality; solenoidal component is structurally invisible.
Score matching errors are not sufficient for measuring diffusion model quality.
problem The L2 score matching error is not a reliable measure of diffusion model performance. method Decomposed score errors into gradient and solenoidal components and analyzed their geometric properties.
result Only the gradient component of the score error affects the marginal distributional quality.
This work improves likelihood of score-based diffusion ODEs using high-order denoising score matching.
problem The gap between maximum likelihood and score matching objectives for score-based diffusion ODEs.
method High-order denoising score matching to maximize likelihood.
result Score-based diffusion ODEs achieve better likelihood on synthetic and CIFAR-10 data.
The paper analyzes the error accumulation in a compositional score-based algorithm for SBI.
problem How to effectively combine multiple observations to improve parameter inference.
method Study of the GAUSS algorithm's compositional score and its mean squared error.
result Established an upper bound on the mean squared error of the compositional score.
The study analyzes and mitigates errors in PC-based causal discovery methods.
problem Errors in PC-based causal discovery methods can lead to incorrect graphs.
method The study introduces coherency scores to detect assumption violations and small sample errors in PC-based methods.
result The coherency scores can detect errors that other methods cannot, bridging between global and local error detection.
Reconstruction error is a prevalent score used to identify anomalous samples when data are modeled by generative models, such as (variational) auto-encoders or generative adversarial networks. This score relies on the assumption that normal samples are located on a manifold and all anomalous samples are located outside…
Improves model calibration for deep neural networks using proper scores.
problem Calibration errors in deep neural networks are often biased and inconsistent.
method Introduces proper calibration errors related to proper scores.
result Demonstrates the superiority of proper scores over common estimators.
Study improves denoising score matching under relaxed manifold assumptions.
problem Improving denoising score matching under relaxed manifold assumptions.
method Model density with nonparametric Gaussian mixtures, relax manifold assumption, derive non-asymptotic bounds.
result Non-asymptotic bounds on approximation and generalization errors, rates of convergence determined by intrinsic dimension.
The paper reconciles two conflicting fairness criteria in algorithmic risk scores.
problem How to reconcile calibration and equal error rates in algorithmic risk scores.
method Derive necessary and sufficient conditions for existence of calibrated scores achieving equal error rates, then present an algorithm to find the most accurate score subject to both criteria.
result The method can eliminate error disparities while maintaining calibration and improve profit in credit lending.
Combining diffusion models with Langevin dynamics improves posterior sampling efficiency.
problem Sampling from noisy posterior distributions efficiently.
method Annealed Langevin dynamics combined with diffusion models.
result Achieves posterior sampling in polynomial time with a weaker score error bound.
Estimates proper calibration errors and refinement terms in probabilistic predictions.
problem Lack of a general estimator for proper calibration errors and refinement terms with known statistical properties.
method Proposes a method for consistent, asymptotically unbiased estimation of proper calibration errors and refinement terms.
result Proves the relation between refinement and f-divergences, implying information monotonicity in neural networks.
New analysis improves convergence guarantees for diffusion-based samplers in Wasserstein distance.
problem Improving convergence guarantees for diffusion-based generative models.
method Simple framework to analyze discretization, initialization, and score estimation errors.
result First Wasserstein convergence bound for the Heun sampler and improved results for Euler sampler.
SCORE resolves the robustness vs accuracy trade-off by redefining robust error.
problem The inherent trade-off between robustness and accuracy in adversarial training.
method SCORE defines local equivariance as the ideal robust behavior, leading to a new robust error metric.
result SCORE reconciles robustness and accuracy, improving model performance on RobustBench.
New characterization limits sampling with inexact scores.
problem Limiting sampling with inexact scores for unbiased results.
method Characterized types of inexact score oracle access.
result Weaker error assumptions rule out tractability of unbiased sampling.
A novel score-based method solves high-dimensional Fokker-Planck equations with improved accuracy and speed.
problem High-dimensional Fokker-Planck equations suffer from the curse of dimensionality, leading to numerical errors and slow sampling.
method Score-based Physics-Informed Neural Networks (PINNs) that fit the score function in SDEs, using three methods: Score Matching, Sliced Score Matching, and Score-PINN.
result The score-based method outperforms traditional Monte Carlo and vanilla PINNs in high-dimensional settings, offering faster sampling and reduced errors.
Paper analyzes stability and forgetting in score-based generative models.
problem Understanding the stability and long-time behavior of generative models.
method Quantitative bounds on sampling error using stability and forgetting properties of the Markov chain.
result Provides practical consequences of stability and contraction mechanism in sampling.
This work analyzes SGD for SGMs, providing convergence rates and error bounds.
problem Optimization dynamics of SGMs trained with stochastic gradients.
method Non-convex convergence rate analysis and Neural Tangent Kernel analysis.
result Theoretical insights into SGD convergence and error bounds for SGMs.
New analysis shows scores learn data manifolds better than distributions.
problem Learning the full distribution vs. just the data manifold.
method Novel analysis of scores in the small-σ regime.
result Scores learn data manifold information Θ(σ−2) stronger than distribution information. The paper analyzes the probabilistic structure of DDPMs and bounds their sampling error.
problem Understanding and controlling errors in discrete-time DDPMs.
method Structural analysis of score functions, Schrödinger's problem, and FBSDEs.
result Explicit upper bound for total variation distance between sampling and target distributions.
Adaptive method improves prediction intervals with global coverage guarantees and local error distribution.
problem Global coverage guarantees of conformal regression are often violated by local error distributions.
method Adaptive Conformal Regression with Jackknife+ Rescaled Scores
result Improves local coverage without sacrificing global coverage, especially in low-data regimes.
Method estimates LLM error rates using Pareto optimization.
problem Quantifying error rates in text-generating models.
method Pareto optimization for generating risk scores.
result Risk scores correlate well with true error rates.
The paper argues for using Neyman orthogonal score for balancing in debiased machine learning.
problem Debiased machine learning requires a proper approach to balance covariates.
method The paper advocates for using Riesz regression with basis functions of X for balancing.
result Covariate balancing is only valid when the score-relevant regression error is a function of covariates alone.
This paper analyzes error bounds for biased SMC samplers in conditional sampling.
problem Analyzing error bounds for biased SMC samplers in conditional sampling.
method Develops a non-asymptotic error analysis for SMC samplers with biased mutation kernels.
result Derives the first non-asymptotic error bound for conditional sampling with score-based diffusion models.
A novel framework quantifies uncertainty using proper scores for various tasks.
problem Uncertainty quantification in machine learning for reliable applications.
method Proposes a general framework based on proper scores for epistemic, aleatoric uncertainty, and model calibration.
result Achieves state-of-the-art uncertainty estimation for large language models and generative models.
New method generates novel samples from closed-form diffusion models.
problem Closed-form SGMs memorize training data and cannot generate novel samples.
method Explicitly smooth closed-form score, use nearest-neighbor estimator.
result Efficient method generates novel samples without training.
New findings show score matching's accuracy doesn't ensure numerical stability in diffusion sampling.
problem Numerical stability issues in diffusion sampling despite small forward-marginal error.
method Constructing a smooth score field with arbitrarily small forward-marginal L2 error, showing nonexplosive behavior and moments of every order. result Euler--Maruyama discretizations can converge in probability even when moments diverge, demonstrating failure of weak convergence.
SGMs are robust to practical errors via uncertainty quantification.
problem Robustness of SGMs to practical implementation errors.
method Wasserstein uncertainty propagation (WUP) theorem and Bernstein estimates.
result SGMs are provably robust to multiple sources of error.
The paper explores various forms of calibration scores and their implications for fairness.
problem The evaluation of probabilistic predictions through calibration.
method The authors organize three grouping choices and one agglomeration of group errors, providing a framework for comparing and creating new calibration scores.
result The study demonstrates that appropriate choices of grouping can provide notions of (sub-)group or individual fairness.
Score-based diffusion models achieve optimal error bounds under non-parametric assumptions.
problem Improving the minimax optimality of score-based diffusion models.
method Kernel-based score estimation and early stopping strategy.
result Achieves minimax optimal error bounds under sub-Gaussian and Sobolev space assumptions.
Synthetic data augmentation can improve imbalanced classification metrics.
problem Improving imbalanced classification metrics
method Developing a framework for analyzing the effects of synthetic data augmentation on score-based classification
result Augmentation can improve AUROC, AUPRC, balanced accuracy, and F1 score
PredictaBoard benchmarks LLM score predictors to assess their ability to anticipate errors.
problem Inconsistent performance of LLMs in common sense reasoning tasks.
method Collaborative benchmarking framework evaluating pairs of LLMs and assessors using rejection rate at different tolerance errors.
result Highlights the need to evaluate predictability alongside performance for safer AI systems.
Post-calibration improves the accuracy of causal effect estimation.
problem Improperly calibrated propensity scores lead to inaccurate causal effect estimation.
method Performed a simulation study to assess the impact of post-calibration on causal effect estimation.
result Post-calibration reduces the error in estimating the average treatment effect, especially for expressive uncalibrated statistical estimators.
SCoRE provides risk control for selective prediction models.
problem Enforcing strict error control in selective prediction models.
method SCoRE framework based on conformal inference and hypothesis testing.
result SCoRE offers binary trust decisions with finite-sample error control.
A new method approximates the exact posterior score for diffusion models.
problem Training-free guidance of diffusion models for image restoration and inverse problems.
method Presented a novel expression for the exact posterior score, leveraging it to compute step sizes on the fly.
result Demonstrated competitive performance with fewer time steps compared to state-of-the-art techniques.
Improved GEC models use scored data from large pretraining to outperform.
problem Addressing data sparsity in Grammatical Error Correction.
method Derive example-level scores from a smaller, higher-quality dataset and incorporate delta-log-perplexity into training schedules.
result Models trained on scored data achieve state-of-the-art results.
Score-based models improve diffuse optical tomography accuracy.
problem Improving accuracy in diffuse optical tomography with uncertainty quantification.
method Score-based diffusion models with a mixed score function to prevent overfitting.
result Data-driven prior distribution results in posterior samples with low variance and centred around the ground truth.
INNs produce interval-valued uncertainty scores for DNNs.
problem Uncertainty quantification in deep neural networks.
method Data-driven interval propagating network using interval arithmetic.
result INNs produce sensible lower and upper bounds for prediction error.
A novel diffusion method for Bayesian posterior sampling with theoretical guarantees.
problem Efficiently sampling from complex posterior distributions in Bayesian inversion.
method Diffusion-based posterior sampling using Langevin dynamics and PnP framework.
result The method converges even for multi-modal posterior distributions with theoretical error bounds.
A network may have weak signals and severe degree heterogeneity, and may be very sparse in one occurrence but very dense in another. SCORE (Jin, 2015) is a recent approach to network community detection. It accommodates severe degree heterogeneity and is adaptive to different levels of sparsity, but its performance for…
TeLeS improves ASR confidence estimation by considering temporal alignment and lexical errors.
problem Inaccurate confidence scores from E2E ASR models, especially for overconfident predictions.
method Proposes TeLeS, a novel confidence score that considers temporal alignment and lexical errors, and uses shrinkage loss to handle data imbalance.
result TeLeS generalizes well across different languages and ASR models, leading to significant WER reduction.
Adaptive learning of SPDE solutions using score-based diffusion models.
problem Model errors and reduced accuracy in SPDE solutions due to incomplete physical knowledge and environmental variability.
method Score-based diffusion models with recursive Bayesian inference, incorporating simulation data and observational information.
result Accuracy and robustness of the proposed method demonstrated on benchmark SPDEs.
Novel method for nonlinear data assimilation using Langevin sampling.
problem Nonlinear data assimilation challenges in Bayesian filtering.
method Score-based sequential Langevin sampling (SSLS) with dynamic models and annealing.
result Asymptotic stability and error bounds for local posterior sampling.
We present a new method for computing ASR word confidences that effectively mitigates the effect of ASR errors for diverse downstream applications, improves the word error rate of the 1-best result, and allows better comparison of scores across different models. We propose 1) a new method for modeling word confidence u…
In this paper, we consider a statistical problem of learning a linear model from noisy samples. Existing work has focused on approximating the least squares solution by using leverage-based scores as an importance sampling distribution. However, no finite sample statistical guarantees and no computationally efficient o…
Designing deterministic denominators for SGLD stabilizes large drifts.
problem Stabilizing large drifts in SGLD
method Using state-dependent envelopes and empirical quantiles for activation thresholds
result Proxy-quantile denominators are close to oracle-score behavior and improve deterministic taming choices
The paper investigates how calibrating propensity scores improves DML estimates of average treatment effects.
problem Improving the accuracy of DML estimates in finite samples.
method Propensity score calibration within the Double/debiased machine learning framework.
result Calibrating propensity scores reduces the root mean squared error of DML estimates of average treatment effects in finite samples.