This research examines rare spurious correlations in neural networks and their impact on accuracy and privacy.
problem Rare spurious correlations in neural networks and their privacy risks.
method Introducing spurious patterns correlated with a fixed class to a few training examples, analyzing ℓ2 regularization and Gaussian noise. result Rare spurious correlations can significantly impact neural network accuracy and privacy, and specific mitigation methods can be effective.
Last layer retraining improves robustness to spurious correlations without high computational costs.
problem Neural networks can rely on spurious features like backgrounds for predictions.
method Simple last layer retraining on large models.
result Last layer retraining matches or outperforms state-of-the-art approaches on spurious correlation benchmarks.
New rationalization method avoids spurious correlations.
problem Spurious correlations in rationalization criteria.
method Invariant rationalization using game theory constraints.
result Rationales generalize better and align with human judgments.
Online algorithm detects and removes spurious features for strong generalization.
problem Predictive models learn spurious correlations from incomplete data.
method Online continual learning algorithm that detects and removes spurious features.
result Algorithm converges to solutions with strong generalization.
The paper argues that normalized mutual information is biased in clustering and community detection.
problem Bias in normalized mutual information for clustering and community detection.
method Introducing a modified version of mutual information to correct for information content and spurious dependence.
result The modified mutual information leads to different conclusions about which algorithms are best for community detection.
Paper tackles group robustness with partially labeled data.
problem Learning invariant representations from datasets with spurious correlations.
method Constructs a constraint set and derives a high probability bound for group assignment. Proposes an optimization algorithm for worst-off group assignments.
result Improvements in minority group's performance while preserving overall accuracy.
Deep models learn spurious features correlated with target, but can still perform well.
problem Spurious correlations in feature learning.
method Empirical risk minimization and specialized group robustness training.
result Feature representations learned by ERM are competitive with specialized methods.
Unified framework for fairness, robustness, and distribution shifts.
problem Diverse failure modes of machine learning systems.
method Formalizes biases as violations of conditional independence and proves equivalence conditions.
result Equivalent effects of biases in different failure modes under specific conditions.
Removing spurious features can hurt model accuracy and disproportionately affect different groups.
problem Interference from spurious features in robust model performance across different groups.
method Characterization and analysis of spurious feature removal in noiseless overparameterized linear regression.
result Removal of spurious features can decrease accuracy and disproportionately affect different groups, even in balanced datasets.
SGD quickly learns a spurious XOR feature before the signal feature, revealing learning dynamics.
problem Over-reliance on spurious correlations in neural networks trained by SGD.
method Theoretical analysis of SGD on two-layer ReLU networks trained on XOR data.
result SGD learns the spurious feature first and exponentially fast, dominating the signal feature.
The paper tackles spurious correlations in machine learning models and introduces counterfactual invariance.
problem Spurious correlations in machine learning models that depend on irrelevant parts of input data.
method The paper uses causal inference to stress test models and introduces counterfactual invariance as a formal requirement.
result Counterfactual invariance is a requirement for models to be robust to irrelevant perturbations in input data.
Mitigates spurious correlations without bias labels.
problem Spurious correlations bias model performance.
method Introduces a novel training objective and debiasing method DPR.
result DPR achieves state-of-the-art performance.
Complex-valued neural networks avoid spurious local minima.
problem Finding spurious local minima in neural networks.
method Proved no spurious local minima for shallow complex neural networks with quadratic activations.
result Complex-valued weights eliminate spurious local minima in neural networks.
CLIP models robustness to spurious features is re-evaluated using a new dataset.
problem Existing robustness tests of CLIP models may not fully reflect their performance on spurious features.
method Crafted a new dataset (CounterAnimal) to reveal CLIP models' reliance on realistic spurious features.
result CLIP models are robust to spurious features learned from their training data, not ImageNet.
Study identifies and analyzes spurious correlations in data-driven models.
problem Spurious correlations in data-driven models are unreliable and hard to detect.
method Collect and analyze synthetic datasets generated from causal graphs to investigate spurious correlations.
result Patterns connecting spurious correlation hypotheses and model design choices were observed.
Self-training avoids spurious features in domain adaptation.
problem Domain shift with large differences between source and target domains.
method Entropy minimization on unlabeled target data, initialized with a source classifier.
result Entropy minimization avoids using spurious features in large domain shifts.
Algorithm removes spurious concepts from neural network representations without harming task performance.
problem Spurious correlations hinder neural network out-of-distribution generalization.
method Iterative algorithm that identifies two orthogonal subspaces in neural network representation.
result Algorithm outperforms existing methods on computer vision and natural language processing benchmarks.
GIB improves neural network generalization by dynamically selecting task-relevant features across different sequential environments.
problem Poor generalization of deep neural networks to unseen environments.
method Proposes a gated information bottleneck (GIB) approach that dynamically drops spurious correlations and selects task-relevant features.
result GIB outperforms other IB approaches in adversarial robustness and OOD detection.
The paper explores grokking in various models, revealing it's not limited to neural networks.
problem Understanding grokking in different models and its mechanism.
method Empirical exploration of grokking in neural networks, Gaussian processes, linear regression, and Bayesian neural networks. Inducing grokking through spurious dimensions.
result Grokking occurs in non-neural architectures, indicating it's not model-specific.
Current OOD benchmarks overestimate model robustness to spurious correlations.
problem Spurious correlations degrade OOD performance, but benchmarks show the opposite.
method Analyze OOD datasets for spurious correlations and derive conditions for robustness.
result Current OOD benchmarks are misspecified and overestimate model robustness.
NURD improves model performance by distilling representations independent of nuisance variables.
problem Models trained under spurious correlations may fail on data with different nuisance-label relationships.
method Developed Nuisance-Randomized Distillation (NURD) to find representations independent of nuisance variables.
result NURD finds representations that perform better regardless of nuisance-label relationships.
Mitigates biases in reward models using variational inference.
problem Spurious correlations in reward models that align large language models with human preferences.
method Formulates data-generating process, identifies non-spurious latent variables, and uses variational inference to recover them.
result Effective mitigation of spurious correlation issues, yielding more robust reward models.
AFR simplifies reducing reliance on spurious features, improving model performance.
problem Reducing reliance on spurious features for out-of-distribution generalization.
method Automatic Feature Reweighting (AFR) updates the model with a weighted loss.
result AFR improves model performance on benchmarks with minimal compute.
This paper explores loss landscapes of sparse neural networks, finding unique characteristics compared to dense networks.
problem Understanding the loss landscape of sparse neural networks, especially one-hidden-layer networks.
method Analyzes sparse networks with dense and sparse final layers, focusing on linear and non-linear models.
result Sparse networks can have no spurious valleys under certain conditions, but spurious valleys and minima can exist for wide sparse networks.
Riemannian gradient descent escapes some spurious critical points on low-rank matrix manifold.
problem Spurious critical points on the boundary of low-rank matrix manifold.
method Riemannian gradient descent with dynamical low-rank approximation and rescaled gradient flow.
result Riemannian gradient descent escapes some spurious critical points on the boundary of the manifold.
SFB uses stable features to adapt unstable ones for better performance.
problem Improving classifier performance on out-of-distribution data by leveraging stable features.
method SFB learns a predictor that separates stable and unstable features, then adapts unstable predictions using stable predictions.
result SFB can learn an asymptotically-optimal predictor without test-domain labels.
JTT improves model worst-group accuracy without group annotations.
problem Low worst-group accuracy in standard ERM models with spurious correlations.
method Two-stage approach: first ERM, then upweight misclassified examples.
result JTT closes 75% of the gap in worst-group accuracy compared to group DRO.
Logit correction improves model performance by correcting spurious correlations.
problem Spurious correlations lead to poor model performance during inference.
method Proposes logit correction (LC) loss to mitigate spurious correlations.
result LC loss outperforms state-of-the-art solutions by 5.5% absolute improvement.
Develops tools to decompose spurious variations in causal models.
problem Understanding and decomposing spurious variations in causal relationships.
method Formal tools for decomposing spurious effects in Markovian and Semi-Markovian models.
result First results on non-parametric decomposition of spurious effects and sufficient conditions for identification.
Gradient flow in phase retrieval escapes spurious minima with high probability.
problem Understanding gradient-based optimization in high-dimensional non-convex functions.
method Analytical and numerical study of gradient dynamics in phase retrieval.
result Gradient flow avoids spurious minima by drifting along unstable directions.
The paper analyzes how deep models memorize spurious features.
problem Understanding how deep models memorize spurious features in training data.
method Characterizes spurious feature memorization via model stability and feature alignment.
result Memorization of spurious features weakens as generalization capability increases.
New method prevents classifiers from relying on spurious correlations.
problem Group invariant learning fails to prevent classifiers from depending on spurious correlations.
method Statistical independence tests to construct groups and reweight samples by group label proportion.
result New method significantly outperforms existing group invariant learning methods in generalizing to spurious correlation shifts.
Neural networks provide a rich class of high-dimensional, non-convex optimization problems. Despite their non-convexity, gradient-descent methods often successfully optimize these models. This has motivated a recent spur in research attempting to characterize properties of their loss surface that may explain such succe…
Boosts causal discovery by dynamically reweighting samples to learn better DAGs.
problem Overfitting to easier-to-fit samples and violating homogeneity assumptions in causal discovery.
method Adaptive sample reweighting via ReScore function to upweight and downweight samples based on fitting quality.
result Consistent and significant boosts in structure learning performance on synthetic and real-world datasets.
CGRL improves graph neural networks' OOD generalization by blocking spurious correlations.
problem Graph Neural Networks struggle with out-of-distribution data due to learning spurious correlations.
method Formulates a causal graph, uses backdoor adjustment, and introduces a loss replacement strategy.
result Significantly improves OOD generalization of GNNs, stabilizing mutual information learning.
DMs emerge from DenseAMs, transitioning from memorization to generalization.
problem Hindered memory retrieval in DenseAMs due to spurious states.
method Examined diffusion models through the lens of DenseAMs, focusing on their generative process.
result Identified a critical phase in DMs transitioning from memorization to generalization.
The FastICA algorithm is one of the most popular iterative algorithms in the domain of linear independent component analysis. Despite its success, it is observed that FastICA occasionally yields outcomes that do not correspond to any true solutions (known as demixing vectors) of the ICA problem. These outcomes are comm…
LaCIM avoids spurious correlation by modeling latent causal factors.
problem Avoiding spurious correlation in supervised learning.
method Introducing latent variables for causal prediction and optimizing over latent space.
result Improved interpretability, robustness, and prediction power on OOD scenarios.
Invariant Causal Set Covering Machines avoid spurious associations.
problem Learning algorithms for rule-based models are vulnerable to spurious associations.
method Building on invariant causal prediction, propose Invariant Causal Set Covering Machines for conjunctions/disjunctions of binary-valued rules.
result The method can identify causal parents of a variable of interest in polynomial time.
We show that for any convex differentiable loss, a deep linear network has no spurious local minima as long as it is true for the two layer case. This reduction greatly simplifies the study on the existence of spurious local minima in deep linear networks. When applied to the quadratic loss, our result immediately impl…
GRASP removes spurious correlations in fine-tuned models, improving task performance and reducing bias.
problem Fine-tuned models can latch onto spurious correlations, leading to bias and reduced generalization.
method GRASP identifies and removes spurious correlations from model weights without removing latent factors.
result GRASP significantly reduces bias and improves task performance in various fine-tuning tasks.
The study identifies spurious correlations in high-dimensional regression and quantifies their impact.
problem Spurious correlations in high-dimensional regression models.
method Statistical characterization of spurious correlations, quantifying their amount via ridge regularization.
result The value of regularization strength that minimizes test loss is in an interval where spurious correlations increase.
When the linear measurements of an instance of low-rank matrix recovery satisfy a restricted isometry property (RIP)---i.e. they are approximately norm-preserving---the problem is known to contain no spurious local minima, so exact recovery is guaranteed. In this paper, we show that moderate RIP is not enough to elimin…
Study on W2S generalization with spurious correlations, proposing remedies.
problem Understanding and improving W2S generalization with spurious correlations.
method Theoretical analysis and algorithmic remedies for W2S fine-tuning.
result W2S always happens with sufficient pseudolabels when group fractions match, but may fail otherwise.
Large learning rates enhance model robustness and compressibility.
problem Achieving robustness and resource-efficiency in machine learning models.
method Identifying and utilizing large learning rates as a facilitator for robustness and compressibility.
result Large learning rates produce desirable representation properties and compare favorably to other methods.
New findings show Bregman proximal algorithms can get stuck near non-stationary points.
problem Bregman proximal algorithms can get stuck near non-stationary points, misleadingly suggesting convergence.
method Analysis of Bregman proximal algorithms and their behavior near non-stationary points.
result Bregman proximal algorithms can get stuck near spurious stationary points, even in convex problems.
DORA analyzes deep neural networks' internal representations to detect spurious correlations.
problem Detecting spurious correlations in deep neural networks' internal representations.
method DORA uses Extreme-Activation (EA) distance measure to assess representation similarities.
result Identifies internal representations capable of detecting spurious correlations.
Nonconvex matrix recovery is known to contain no spurious local minima under a restricted isometry property (RIP) with a sufficiently small RIP constant δ. If δ is too large, however, then counterexamples containing spurious local minima are known to exist. In this paper, we introduce a proof technique that is capa…