Improved scalability and interpretability in training data attribution.
problem Identifying which training data drives specific behaviors, especially unintended ones.
method Leveraging interpretable structures within the model to attribute model behavior to semantic directions, not individual test examples.
result Simple probe-based attribution methods are first-order approximations of Concept Influence that achieve comparable performance while being over an order-of-magnitude faster.
Improved GP models for fast training and good performance.
problem Training scalable Gaussian process models efficiently.
method Cross-validation and nearest neighbor truncation for scalable GP training.
result Our method offers fast training and excellent predictive performance.
A scalable method for training prediction models in predict-then-optimize
problem Training prediction models in the predict-then-optimize paradigm
method Decision-focused learning pipeline
result Decision quality competitive with state-of-the-art methods while reducing training time
TensorHyper-VQC improves VQC scalability and robustness.
problem Scalability and noise sensitivity in VQC.
method Tensor-train-guided hypernetwork framework.
result TensorHyper-VQC achieves superior performance and robust noise tolerance.
This paper introduces a neural sampler for scalable sampling from complex distributions.
problem Efficiently sampling from high-dimensional un-normalized distributions.
method Neural implicit sampler trained with KL and Fisher divergence methods.
result The neural sampler generates large batches of samples with low computational costs.
This paper shows using classification instead of regression improves deep RL scalability.
problem Challenges in training value functions for large networks in deep RL.
method Used categorical cross-entropy loss instead of mean squared error regression.
result Significant improvements in performance and scalability across various domains.
In this paper, we present a general framework to scale graph autoencoders (AE) and graph variational autoencoders (VAE). This framework leverages graph degeneracy concepts to train models only from a dense subset of nodes instead of using the entire graph. Together with a simple yet effective propagation mechanism, our…
ShadowSync separates background synchronization for scalable distributed training.
problem Reducing synchronization overhead in distributed training for high scalability.
method Separates synchronization from training and runs it in the background.
result Achieves both high throughput and excellent model quality at scale.
Meta-learning interpretable decision trees with synthetic data.
problem Lack of efficient, scalable methods for generating synthetic data for decision tree meta-learning.
method Synthetic generation of near-optimal decision trees using the MetaTree transformer architecture.
result Meta-learning of decision trees achieves performance comparable to real-world data or optimal decision trees, with significant computational cost reduction.
Graphs are ubiquitous real-world data structures, and generative models that approximate distributions over graphs and derive new samples from them have significant importance. Among the known challenges in graph generation tasks, scalability handling of large graphs and datasets is one of the most important for practi…
Bayesian tensor train kernel machine uses Laplace approximation for scalable GP regression.
problem Scalability limitations of Gaussian process regression.
method Bayesian tensor train kernel machine with Laplace approximation and variational inference.
result VI replaces cross-validation and offers up to 65x faster training.
COPML framework securely trains models across multiple data owners without revealing individual data.
problem Privacy-preserving collaborative machine learning with multiple data owners.
method Securely encodes data, distributes computation, performs distributed training.
result Achieves up to 16x speedup in training time while maintaining strong privacy.
Decentralized Gaussian processes for multi-agent learning.
problem Training and prediction in multi-agent systems.
method Decentralized ADMM for GP hyper-parameter training and iterative consensus for prediction.
result Subset of agents can perform predictions using covariance-based nearest neighbor selection.
The practical usage of reinforcement learning agents is often bottlenecked by the duration of training time. To accelerate training, practitioners often turn to distributed reinforcement learning architectures to parallelize and accelerate the training process. However, modern methods for scalable reinforcement learnin…
A scalable ROC-SVM variant reduces training time for imbalanced binary classification.
problem High computational cost of ROC-SVM for imbalanced binary classification.
method Incomplete U-statistics and low-rank kernel approximation.
result Comparable AUC performance with reduced training time.
Scalable Gaussian process models trained with unbiased stochastic ELBO.
problem Training large capacity Gaussian process models on huge datasets.
method Unbiased stochastic variational inference for scalable GPs.
result Accurate inference on large datasets with up to 10 million basis functions.
New algorithm improves Gaussian process hyperparameter tuning for large datasets.
problem Scalable hyperparameter tuning for Gaussian processes on large datasets.
method Estimates smoothness and length-scale parameters in Matern kernel using novel loss functions.
result Improved uncertainty quantification over traditional methods.
We propose a distributed approach to train deep neural networks (DNNs), which has guaranteed convergence theoretically and great scalability empirically: close to 6 times faster on instance of ImageNet data set when run with 6 machines. The proposed scheme is close to optimally scalable in terms of number of machines, …
The paper discusses scalable learning for wireless data-driven systems.
problem Expanding data volume and model complexity limit centralized learning solutions.
method Discusses scalable architecture and local learning strategies.
result Promising research directions in scalable data-driven wireless communications.
Scalable feature selection improves GBDT model training speed.
problem Efficiently selecting features for GBDT models in high-dimensional settings.
method Developed a scalable forward feature selection method using group testing.
result Significant speedups in training time with competitive model performance.
NeuroMAS treats multi-agent systems as neural networks for scalable, trainable coordination.
problem Designing multi-agent systems as hand-designed workflows is inefficient and inflexible.
method NeuroMAS treats multi-agent systems as a neural network architecture with reinforcement learning for scalable coordination.
result NeuroMAS improves significantly over multi-agent baselines and can be scaled progressively.
FinRL-Podracer accelerates DRL trading strategies in finance with high performance and scalability.
problem Challenges in applying deep reinforcement learning to finance trading models.
method Proposes an RLOps framework and high-performance cloud solution for DRL trading.
result FinRL-Podracer outperforms existing DRL libraries by 12-35% in annual return, 0.1-0.6 in Sharpe ratio, and 3-7 times in training time.
The Graph Convolutional Network (GCN) model and its variants are powerful graph embedding tools for facilitating classification and clustering on graphs. However, a major challenge is to reduce the complexity of layered GCNs and make them parallelizable and scalable on very large graphs -- state-of the art techniques a…
L2G2G improves graph autoencoder accuracy without sacrificing scalability.
problem Scalability issues in graph autoencoders for large networks.
method Dynamic synchronization of latent node representations in each epoch.
result L2G2G achieves higher accuracy than standard Local2Global approach on large and dense networks.
New method for efficient sketching of gradients and Hessians.
problem Memory constraints in training machine learning models.
method A novel framework for scalable gradient and HVP sketching tailored for modern hardware.
result Theoretical guarantees and practical applications in training data attribution and Hessian spectrum analysis.
New algorithm reduces bias in trained models, near-optimal performance proven.
problem Reduction of bias in trained machine learning models.
method Scalable post-processing algorithm for debiasing trained models, including deep neural networks (DNNs).
result Proven to be near-optimal by bounding its excess Bayes risk.
BNAS improves neural architecture search with a scalable, fast, and efficient approach.
problem Efficiently searching for optimal neural architectures with high performance and low training time.
method Designing a broad scalable architecture (BCNN) with reinforcement learning and parameter sharing, and developing two variants.
result Significantly reduces training time and achieves state-of-the-art performance on CIFAR-10 and ImageNet.
A scalable deep learning framework accelerates training of large neural networks for solving 3D Poisson equations.
problem Training large-scale neural networks for solving complex PDEs efficiently.
method Combines multigrid techniques with distributed deep learning to accelerate training.
result Solves 3D Poisson equations up to 512x512x512 resolution efficiently.
FSBM improves matching efficiency with minimal supervision.
problem Scalability vs. minimal supervision in matching frameworks.
method FSBM uses a small portion of pre-aligned pairs as state feedback to guide non-coupled samples.
result FSBM accelerates training and enhances generalization.
EduQG generates better educational questions by pre-training on scientific text.
problem Improving the quality of educational questions for scalable self-assessment.
method Adapting a large language model for educational question generation, pre-trained on scientific text.
result EduQG produces superior educational questions compared to baseline models.
We introduce a scalable measure of curvature for analyzing training dynamics of large language models.
problem Analyzing the training dynamics of large language models due to high computational cost of measuring Hessian sharpness.
method We introduce critical sharpness and relative critical sharpness as computationally efficient measures capturing Hessian sharpness phenomena.
result We provide the first demonstration of sharpness phenomena at scale up to 7B parameters.
Synchronized stochastic gradient descent (SGD) optimizers with data parallelism are widely used in training large-scale deep neural networks. Although using larger mini-batch sizes can improve the system scalability by reducing the communication-to-computation ratio, it may hurt the generalization ability of the models…
Dreamer 4 learns Minecraft tasks from videos alone.
problem Accurately predicting object interactions in complex environments.
method Reinforcement learning inside a fast, accurate world model.
result Dreamer 4 outperforms previous models in Minecraft, learning from only offline data.
New SMC method for pBNNs improves scalability and predictive performance.
problem Training pBNNs with high-dimensional stochastic parameters.
method Gradient-based proposals within SMC samplers.
result New method outperforms state-of-the-art in predictive performance and training time.
New method for scalable set encoding with unbiased gradient approximation.
problem Limited expressive power and large set training issues in set functions.
method Universally MBC (UMBC) class of set functions and efficient MBC training algorithm.
result Unbiased approximation of full set gradient with constant memory overhead.
Quantum machine learning: Adiabatic quantum SVM outperforms classical methods.
problem Training support vector machines efficiently on large datasets.
method Adiabatic quantum computing for SVM training.
result Quantum approach outperforms classical methods in accuracy and scalability.
SCALLOP improves likelihood flow maps for efficient Boltzmann generation.
problem Efficient estimation of model likelihood in flow-based generative models.
method SCALLOP introduces a Hutchinson-free likelihood distillation objective for scalable flow-based models.
result SCALLOP achieves up to 10x inference speedup while improving performance.
New method improves sampling from noisy energy models.
problem Training and sampling challenges in Energy-Based Models.
method Pseudo-Gibbs sampling with moment matching.
result Effective sampling from clean model using noisy DSM-trained model.
Inference in Gaussian process (GP) models is computationally challenging for large data, and often difficult to approximate with a small number of inducing points. We explore an alternative approximation that employs stochastic inference networks for a flexible inference. Unfortunately, for such networks, minibatch tra…
As a non-parametric Bayesian model which produces informative predictive distribution, Gaussian process (GP) has been widely used in various fields, like regression, classification and optimization. The cubic complexity of standard GP however leads to poor scalability, which poses challenges in the era of big data. Hen…
Data loading can dominate deep neural network training time on large-scale systems. We present a comprehensive study on accelerating data loading performance in large-scale distributed training. We first identify performance and scalability issues in current data loading implementations. We then propose optimizations t…
In stochastic optimization, using large batch sizes during training can leverage parallel resources to produce faster wall-clock training times per training epoch. However, for both training loss and testing error, recent results analyzing large batch Stochastic Gradient Descent (SGD) have found sharp diminishing retur…
LeJEPA provides a scalable, theory-driven approach to self-supervised learning.
problem Lack of practical guidance and theory in JEPAs.
method Identified optimal Gaussian distribution and introduced SIGReg objective.
result LeJEPA achieves state-of-the-art performance with minimal hyperparameters and heuristics.
Securely trains regression models with secret sharing for data collaboration.
problem Balancing data collaboration for technological improvements with security concerns.
method Secret sharing scheme for scalable and efficient secure multiparty training.
result Scalable and efficient protocols for training linear and logistic regression models.
SDE Matching eliminates simulation for training Latent SDEs, achieving similar performance.
problem Training Latent SDEs with adjoint sensitivity methods is computationally expensive and limited.
method SDE Matching, inspired by Score- and Flow Matching, eliminates simulation for training Latent SDEs.
result SDE Matching achieves performance comparable to adjoint sensitivity methods while reducing computational complexity.
NHC learns scalable algorithmic solutions from diverse tasks.
problem Neural networks struggle to learn algorithmic strategies.
method Memory-augmented network with abstraction mechanism and evolutionary training.
result Reliable learning of robust and scalable algorithmic solutions.
Framework handles both exchangeable and non-exchangeable event sequences without tuning.
problem Handling both exchangeable and non-exchangeable event sequences efficiently.
method Parametric Hawkes-process-inspired conditional probability mass function with variational inference.
result Competitive computational and predictive performance against state-of-the-art methods.
Paper presents a new method to learn unnormalized models efficiently.
problem Scalability issues in score matching for flexible unnormalized models.
method Connects learning objectives to Wasserstein gradient flows for scalability.
result Demonstrates improved learning of unnormalized models on manifolds.