arXiv research

Single-step samplers generate high-quality samples efficiently.

Paper tackles catastrophic overfitting in single-step adversarial training.

Unified model improves sampling speed and quality.

This paper establishes a theoretical foundation for consistency training in diffusion models.

Adversarial examples have become one of the largest challenges that machine learning models, especially neural network classifiers, face. These adversarial examples break the assumption of attack-free scenario and fool state-of-the-art (SOTA) classifiers with insignificant perturbations to human. So far, researchers ac…

CARV reduces compute cost for downstream pipelines using diffusion models.

Efficient regularization mitigates catastrophic overfitting in single-step adversarial training.

A new method reduces compounding errors in model-based reinforcement learning.

A new method reduces adversarial training time without overfitting.

We present an extension of our Molecular Transformer architecture combined with a hyper-graph exploration strategy for automatic retrosynthesis route planning without human intervention. The single-step retrosynthetic model sets a new state of the art for predicting reactants as well as reagents, solvents and catalysts…

Adversarial examples are malicious inputs designed to fool machine learning models. They often transfer from one model to another, allowing attackers to mount black box attacks without knowledge of the target model's parameters. Adversarial training is the process of explicitly training a model on adversarial examples,…

Lazy neural networks are vulnerable to adversarial attacks.

M-FFF generates data on manifolds with fast sampling.

A new sampler for complex discrete distributions efficiently updates all variables in parallel.

A novel optimisation framework through quadratic nonlinear projection is introduced for credit portfolio when the portfolio risk is measured by Conditional Value-at-Risk (CVaR). The whole optimisation procedure to search toward the optimal portfolio state is conducted by a series of single-step optimisations under the …

Improved chemical reaction prediction using augmented NLP models.

New algorithm improves sampling from constrained spaces.

New sampling method improves accuracy for constrained spaces.

Hadamard Wirtinger Flow recovers sparse signals from fewer measurements.

New method improves adversarial training efficiency and robustness.

CTM improves diffusion model sampling quality with efficient ODE traversal.

This paper deals with dependence across marginally exponentially distributed arrival times, such as default times in financial modeling or inter-failure times in reliability theory. We explore the relationship between dependence and the possibility to sample final multivariate survival in a long time-interval as a sequ…

New framework for modular reinforcement learning reduces sample complexity.

A new algorithm combines SGD and Thompson Sampling for contextual bandits.

The problem of automatic and accurate forecasting of time-series data has always been an interesting challenge for the machine learning and forecasting community. A majority of the real-world time-series problems have non-stationary characteristics that make the understanding of trend and seasonality difficult. Our int…

Efficient algorithm for learning from indirect feedback in complex decision-making scenarios.

W-Flow generates images in one step, faster and better than multi-step methods.

A new approach RA improves stochastic optimization by executing multiple steps between subsample updates.

Unified framework improves meta-learning generalization bounds.

This paper improves the efficiency of generative models by optimizing the straightness of Rectified Flow.

Certified training improves robustness against adversarial attacks.

Adversarial examples are perturbed inputs designed to fool machine learning models. Adversarial training injects such examples into training data to increase robustness. To scale this technique to large datasets, perturbations are crafted using fast single-step methods that maximize a linear approximation of the model'…

Action-bisimulation learns long-horizon controllability for reinforcement learning.

Gradient descent memorizes many Gaussians efficiently.

A new TS-SA method alleviates non-stationarity in TS algorithms for bandits.

Efficiently clusters data sequences with varying distributions.

The goal in network state prediction (NSP) is to classify the global state (label) associated with features embedded in a graph. This graph structure encoding feature relationships is the key distinctive aspect of NSP compared to classical supervised learning. NSP arises in various applications: gene expression samples…

We show that the lack of arbitrage in a model with both fixed and proportional transaction costs is equivalent to the existence of a family of absolutely continuous single-step probability measures, together with an adapted process with values between the bid-ask spreads that satisfies the martingale property with resp…

Paper proposes a new method to speed up diffusion models.

A new method improves stochastic gradient descent for faster and more efficient estimation.

We study an exploration method for model-free RL that generalizes the counter-based exploration bonus methods and takes into account long term exploratory value of actions rather than a single step look-ahead. We propose a model-free RL method that modifies Delayed Q-learning and utilizes the long-term exploration bonu…

By injecting adversarial examples into training data, adversarial training is promising for improving the robustness of deep learning models. However, most existing adversarial training approaches are based on a specific type of adversarial attack. It may not provide sufficiently representative samples from the adversa…

Generative models learn distributions, new method finds inputs matching desired conditional distributions.

Efficient methods for answering complex probabilistic queries in sequential data.

Improves spatio-temporal forecasting by reducing errors between training and inference.

In this work, we describe practical lessons we have learned from successfully using contextual bandits (CBs) to improve key business metrics of the Microsoft Virtual Agent for customer support. While our current use cases focus on single step einforcement learning (RL) and mostly in the domain of natural language proce…

When environmental interaction is expensive, model-based reinforcement learning offers a solution by planning ahead and avoiding costly mistakes. Model-based agents typically learn a single-step transition model. In this paper, we propose a multi-step model that predicts the outcome of an action sequence with variable …

XCS constitutes the most deeply investigated classifier system today. It bears strong potentials and comes with inherent capabilities for mastering a variety of different learning tasks. Besides outstanding successes in various classification and regression tasks, XCS also proved very effective in certain multi-step en…