This work tackles semi-supervised federated learning by reducing model gradient diversity.
problem Improving test accuracy in semi-supervised federated learning with limited labeled data.
method Investigates and compares various design choices including consistency regularization loss, Batch Normalization, and Group Normalization.
result Grouping-based model averaging combined with Group Normalization and consistency regularization loss improves test accuracy.
We design a new algorithm for batch active learning with deep neural network models. Our algorithm, Batch Active learning by Diverse Gradient Embeddings (BADGE), samples groups of points that are disparate and high-magnitude when represented in a hallucinated gradient space, a strategy designed to incorporate both pred…
The paper improves QD policy ensembles using distribution ratio estimators.
problem Training diverse and high-quality reinforcement learning agents.
method Using Stein variational gradient descent and distribution ratio estimators.
result The method generates diverse and high-quality reinforcement learning agents.
New method learns diverse solutions in reinforcement learning without gradient bias.
problem Lack of diverse solutions in reinforcement learning tasks.
method Maximizes state-action-based mutual information directly, using variational lower bound.
result Successfully learns an infinite set of diverse solutions.
Enhances sample diversity in SGMCMC for better uncertainty estimation in BNNs.
problem Limited sample diversity in SGMCMC affects uncertainty estimation and model performance.
method Reparameterizes neural network weights to produce a more diverse set of samples.
result The proposed approach achieves superior performance in image classification tasks, including OOD robustness.
Standard reinforcement learning methods aim to master one way of solving a task whereas there may exist multiple near-optimal policies. Being able to identify this collection of near-optimal policies can allow a domain expert to efficiently explore the space of reasonable solutions. Unfortunately, existing approaches t…
We address the challenge of effective exploration while maintaining good performance in policy gradient methods. As a solution, we propose diverse exploration (DE) via conjugate policies. DE learns and deploys a set of conjugate policies which can be conveniently generated as a byproduct of conjugate gradient descent. …
CWGD measures gradient diversity weighted by curvature, improving SGD convergence.
problem Gradient noise in high-curvature directions is underestimated by standard methods.
method CWGD weights gradient diversity by the inverse square root of the Hessian.
result CWGD-Cosine reduces optimization error by up to 20% compared to standard cosine annealing.
New method shows random, diverse initializations are not essential for deep neural networks.
problem The necessity of random, diverse initializations in deep neural networks.
method Constructed a deep convolutional network with identical features by initializing weights to 0, enabling signal propagation and stable gradients.
result Random, diverse initializations are not necessary for training neural networks.
A new method for diverse Pareto solutions in multi-objective learning.
problem Maximizing diversity while maximizing hypervolume in Pareto solutions.
method Annealed Stein Variational Gradient Descent (SVGD) with diverse gradient directions.
result SVH-MOL achieves superior performance in multi-objective and multi-task learning.
FoRDE uses input gradients to improve neural network ensembles.
problem Improving neural network ensembles for robustness and accuracy.
method Proposes FoRDE, an ensemble learning method based on ParVI, which repels function space by input gradients.
result FoRDE significantly outperforms DEs and other ensemble methods in accuracy and calibration.
The two key players in Generative Adversarial Networks (GANs), the discriminator and generator, are usually parameterized as deep neural networks (DNNs). On many generative tasks, GANs achieve state-of-the-art performance but are often unstable to train and sometimes miss modes. A typical failure mode is the collapse o…
New ACE cost function encourages diversity in neural networks.
problem Training multiple classifiers with controlled diversity.
method Mathematical derivation and gradient control.
result ACE yields better ensemble results than vanilla.
Gradient descent can efficiently learn a target function with diverse and near-orthogonal features.
problem Learning a target function with additive structure and diverse features.
method Gradient descent training of a two-layer neural network.
result A large subset of polynomial target functions can be efficiently learned.
Paper proposes SDRL to improve continual learning with less computational cost.
problem Catastrophic forgetting in continual learning.
method SDRL method that refines gradients from memorized samples to reduce gradient diversity.
result SDRL shows better performance than state-of-the-art methods on multiple benchmark tasks.
Meta-learning improves few-shot land cover classification across diverse regions.
problem Capturing diversity in land cover classification across different geographic regions.
method Model-agnostic meta-learning (MAML) algorithm applied to classification and segmentation tasks.
result Few-shot model adaptation outperforms traditional methods in diverse land cover classification tasks.
New findings on representation learning beyond linear functions.
problem Achieving diversity in representation learning beyond linear prediction functions.
method Analysis of eluder dimension and empirical risks.
result Diversity holds even with nonlinear prediction functions and multiple layers in neural networks.
Entropy minimization has been widely used in unsupervised domain adaptation (UDA). However, existing works reveal that entropy minimization only may result into collapsed trivial solutions. In this paper, we propose to avoid trivial solutions by further introducing diversity maximization. In order to achieve the possib…
Exploration is a key problem in reinforcement learning, since agents can only learn from data they acquire in the environment. With that in mind, maintaining a population of agents is an attractive method, as it allows data be collected with a diverse set of behaviors. This behavioral diversity is often boosted via mul…
Gradient-based meta-learners such as MAML are able to learn a meta-prior from similar tasks to adapt to novel tasks from the same distribution with few gradient updates. One important limitation of such frameworks is that they seek a common initialization shared across the entire task distribution, substantially limiti…
Repulsive deep ensembles improve diversity and Bayesian inference.
problem Challenges in maintaining diversity among ensemble members trained independently.
method Introducing a repulsive term in the update rule of deep ensembles.
result Training dynamics of repulsive ensembles follow a Wasserstein gradient flow of KL divergence with the true posterior.
A new method MixGDA combines mixup and gradient-based data augmentation for SSL.
problem Improving semi-supervised learning performance with limited labeled data.
method Gradient-based Data Augmentation (GDA) combined with mixup methods.
result MixGDA achieves state-of-the-art performance in various SSL benchmarks.
RL enhances LLM planning but introduces spurious solutions and diversity collapse.
problem Theoretical understanding of RL's benefits and limitations in LLM planning.
method Graph-based abstraction, policy gradient, Q-learning, supervised fine-tuning.
result RL's exploration is crucial for generalization, but PG suffers from diversity collapse.
Transformers learn to use induction heads or shortcuts based on data diversity.
problem How data diversity influences the behavior of transformers.
method Gradient-based training of a single-layer transformer on a minimal task.
result Data diversity steers transformers toward induction heads or shortcuts.
Stein variational neural network ensembles improve diversity and uncertainty estimation.
problem Lack of proper Bayesian justification and diversity guarantees in deep neural network ensembles.
method Particle-based inference methods, specifically Stein variational gradient descent (SVGD), operating in weight space, function space, and hybrid settings.
result SVGD methods improve diversity and uncertainty estimation, approaching the true Bayesian posterior more closely.
Framework improves gradient estimation for faster training convergence.
problem Efficiently estimating noisy gradients in stochastic optimization.
method Dynamic adaptive importance sampling combining multiple distributions.
result Adaptively weighted multiple importance sampling yields superior gradient estimates.
The paper proposes blending gradient boosted trees and neural networks for hierarchical time series forecasting.
problem Point and probabilistic forecasting of hierarchical time series.
method A blending methodology of gradient boosted trees and neural networks, with feature engineering and diverse model selection.
result Ranked within the gold medal range in both Accuracy and Uncertainty tracks of the M5 Competition.
We study a mini-batch diversification scheme for stochastic gradient descent (SGD). While classical SGD relies on uniformly sampling data points to form a mini-batch, we propose a non-uniform sampling scheme based on the Determinantal Point Process (DPP). The DPP relies on a similarity measure between data points and g…
ODS improves adversarial attacks by maximizing output diversity.
problem Efficiency and effectiveness of adversarial attacks, especially black-box attacks.
method Output Diversified Sampling (ODS) that maximizes diversity in model outputs.
result ODS reduces the number of queries needed for black-box attacks on ImageNet by a factor of two.
PDNAS optimizes GNN architectures for diverse datasets.
problem Inadequate adaptability and combinatorial search space in GNNs.
method Dual architecture search (micro- and macro-architectures) with gradient-based optimization.
result PDNAS finds deeper GNNs with better performance on diverse datasets.
Paper proposes an online learning method with multi-level adaptivity for diverse loss functions.
problem Online learning with unknown types and curvatures of functions.
method Multi-layer online ensemble approach with gradient variations.
result Achieves improved regret bounds for different types of loss functions.
UCPO improves diversity in reinforcement learning models, maintaining high accuracy.
problem RLVR objectives often lead to diversity collapse, reducing coverage of correct solutions.
method UCPO adds a conditional uniformity penalty to GRPO, redistributing probability mass.
result UCPO improves Pass@K and diversity while maintaining competitive Pass@1 accuracy.
Paper presents a rank-1 approximation method for natural policy gradients in deep RL.
problem Computing natural gradients requires inverting the Fisher Information Matrix, which is computationally expensive.
method Develops a rank-1 approximation to the inverse Fisher Information Matrix for efficient natural policy optimization.
result The rank-1 approximation converges faster and has similar sample complexity to stochastic policy gradient methods.
Standard acquisition functions are sufficient for asynchronous Bayesian optimization.
problem Redundant and repeated queries in asynchronous Bayesian optimization.
method Conceptual analysis and theoretical guarantees of standard acquisitions.
result Standard acquisition functions achieve theoretical guarantees equivalent to Thompson sampling in asynchronous settings.
In recent years, plenty of metrics have been proposed to identify networks that are free of gradient explosion and vanishing. However, due to the diversity of network components and complex serial-parallel hybrid connections in modern DNNs, the evaluation of existing metrics usually requires strong assumptions, complex…
Diverse projection ensembles improve distributional reinforcement learning.
problem Learning the distribution of returns in reinforcement learning.
method Combining multiple projection methods to improve model diversity and exploration.
result Diverse projection ensembles lead to significant performance improvements in exploration tasks.
Batch Normalization (BN) has become a cornerstone of deep learning across diverse architectures, appearing to help optimization as well as generalization. While the idea makes intuitive sense, theoretical analysis of its effectiveness has been lacking. Here theoretical support is provided for one of its conjectured pro…
We introduce Generalized Integrated Gradients (GIG), a formal extension of the Integrated Gradients (IG) (Sundararajan et al., 2017) method for attributing credit to the input variables of a predictive model. GIG improves IG by explaining a broader variety of functions that arise from practical applications of ML in do…
Minimizing the empirical risk is a popular training strategy, but for learning tasks where the data may be noisy or heavy-tailed, one may require many observations in order to generalize well. To achieve better performance under less stringent requirements, we introduce a procedure which constructs a robust approximati…
Gradient oversmoothing and expansion hinder deep GNN training, solved with normalization.
problem Gradient oversmoothing and expansion prevent deep GNN training.
method Proposed normalization method to constrain the Lipschitz bound of each layer.
result Residual GNNs with hundreds of layers can be efficiently trained with the proposed normalization.
Optimizes forecast accuracy and diversity using multi-task deep learning.
problem Forecasting combinations of time series data.
method Multi-task deep learning architecture that selects and combines forecasting models.
result Enhances point forecast accuracy compared to state-of-the-art methods.
Optimising discrete data for a desired characteristic using gradient-based methods involves projecting the data into a continuous latent space and carrying out optimisation in this space. Carrying out global optimisation is difficult as optimisers are likely to follow gradients into regions of the latent space that the…
New algorithm improves dataset distillation with 108% improvement on ImageNet.
problem Improving dataset distillation performance.
method Reparameterization and convexification of implicit gradients (RCIG).
result Establishes new state-of-the-art on various dataset distillation tasks.
Mini-batch gradient descent based methods are the de facto algorithms for training neural network architectures today. We introduce a mini-batch selection strategy based on submodular function maximization. Our novel submodular formulation captures the informativeness of each sample and diversity of the whole subset. W…
Extends gradient-based optimization to spline functions.
problem Limitations of standard differentiable programming methods.
method Derives Jacobian of spline functions and uses it in predictive models.
result Improved performance in various applications.
EGAB algorithms improve online portfolio selection.
problem Online portfolio selection problem.
method Generalized exponentiated gradient (EG) updates with Alpha-Beta divergence regularization.
result EGAB algorithms enhance portfolio performance, especially with transaction costs.
ESS-Flow guides flow models without retraining, using Bayesian inference in source space.
problem Training flow models on paired data for conditional generation or sample production.
method Gradient-free Bayesian inference in source space using Elliptical Slice Sampling.
result Effective in diverse tasks including material design and protein structure prediction.
Neural networks are a powerful class of functions that can be trained with simple gradient descent to achieve state-of-the-art performance on a variety of applications. Despite their practical success, there is a paucity of results that provide theoretical guarantees on why they are so effective. Lying in the center of…