arXiv research

EM Distillation simplifies diffusion models to one-step generators.

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

Unified framework for data-free sampling using Wasserstein gradient flows.

One-step diffusion samplers reduce sampling time and computational costs.

Distributed statistical inference has recently attracted enormous attention. Many existing work focuses on the averaging estimator. We propose a one-step approach to enhance a simple-averaging based distributed estimator. We derive the corresponding asymptotic properties of the newly proposed estimator. We find that th…

A new machine learning method for Bayesian inverse problems in function spaces.

A new one-step method for covariate shift adaptation.

Consistency models generate high-quality samples fast and without iterative sampling.

FastVoiceGrad speeds up VC to one step, matching or surpassing quality.

SMT trains generative models by estimating mixture scores, outperforming existing methods.

Proposes a graph dynamics prior for more accurate relational inference.

FasterVoiceGrad speeds up VC by 6-7x with novel distillation.

VFMs use noise adapters to conditionally generate images in one step.

One-step policy improvement outperforms iterative RL methods on D4RL.

DLM-One speeds up language generation by 500x with continuous models.

The easiness at which adversarial instances can be generated in deep neural networks raises some fundamental questions on their functioning and concerns on their use in critical systems. In this paper, we draw a connection between over-generalization and adversaries: a possible cause of adversaries lies in models desig…

One-step Bellman alignment improves online RL by reducing task mismatch.

In this paper we propose using the principle of boosting to reduce the bias of a random forest prediction in the regression setting. From the original random forest fit we extract the residuals and then fit another random forest to these residuals. We call the sum of these two random forests a \textit{one-step boosted …

We consider a problem of stochastic online learning with general probabilistic graph feedback, where each directed edge in the feedback graph has probability $p_{ij}$. Two cases are covered. (a) The one-step case, where after playing arm $i$ the learner observes a sample reward feedback of arm $j$ with independent prob…

Simpler one-step distributional RL framework for control.

Under a Bayesian framework, we formulate the fully sequential sampling and selection decision in statistical ranking and selection as a stochastic control problem, and derive the associated Bellman equation. Using value function approximation, we derive an approximately optimal allocation policy. We show that this poli…

The paper improves model-based reinforcement learning by using multi-timestep objectives.

LoRA-One uses one-step full gradient to align adapters for efficient large model fine-tuning.

NCT simplifies one-step generator adaptation to new controls.

A novel method for clustering multi-view data using dual representations.

Two BO methods improve reliability optimization for rare failures.

One-step learning in crosspoint memory reduces computation time.

New method uses model comparison signals to improve LLM evaluation accuracy.

Improved knowledge gradient (iKG) outperforms the original KG algorithm in best arm identification problems.

We consider Model-Agnostic Meta-Learning (MAML) methods for Reinforcement Learning (RL) problems, where the goal is to find a policy using data from several tasks represented by Markov Decision Processes (MDPs) that can be updated by one step of stochastic policy gradient for the realized MDP. In particular, using stoc…

Itô maps provide a method for any-step SDE integration.

Efficiently distills pretrained text-to-image models without real data, improving FID and CLIP scores.

IMM generates high-quality samples in few steps with stable training.

The paper approximates supply curves using a one-step basis method.

SiD distills pretrained diffusion models into a fast one-step generator.

The paper develops a theory for one-step Wasserstein-guided models for PDE-induced measures.

The Teacher Forcing algorithm trains recurrent networks by supplying observed sequence values as inputs during training and using the network's own one-step-ahead predictions to do multi-step sampling. We introduce the Professor Forcing algorithm, which uses adversarial domain adaptation to encourage the dynamics of th…

New method trains neural samplers to sample from multi-modal distributions efficiently.

Model-based reinforcement learning is an appealing framework for creating agents that learn, plan, and act in sequential environments. Model-based algorithms typically involve learning a transition model that takes a state and an action and outputs the next state---a one-step model. This model can be composed with itse…

A new method learns straight trajectories in one step for optimal flow matching.

Dynamic risk measures follow law invariance principles over time.

Automatic machine learning-based detectors of various psychological and social phenomena (e.g., emotion, stress, engagement) have great potential to advance basic science. However, when a detector $d$ is trained to approximate an existing measurement tool (e.g., a questionnaire, observation protocol), then care must be…

The performance of policy gradient methods is sensitive to hyperparameter settings that must be tuned for any new application. Widely used grid search methods for tuning hyperparameters are sample inefficient and computationally expensive. More advanced methods like Population Based Training that learn optimal schedule…

The paper introduces a multi-step loss function to improve model-based reinforcement learning.

Robust Transformer-Based One-Step Stock Index Forecasting via Shifted Data Augmentation

New Q-learning method achieves optimal sample complexity for average-reward problems.

Injecting adversarial examples during training, known as adversarial training, can improve robustness against one-step attacks, but not for unknown iterative attacks. To address this challenge, we first show iteratively generated adversarial images easily transfer between networks trained with the same strategy. Inspir…

New insights into neural network feature learning through multi-step gradient descent.