Related papers: AESOP: Adversarial Execution-path Selection to Ove…

Denoising Diffusion Probabilistic Models (DDPMs) have established a new state-of-the-art in generative image synthesis, yet their deployment is hindered by significant computational overhead during inference, often requiring up to 1,000…

The term GreenAI refers to a novel approach to Deep Learning, that is more aware of the ecological impact and the computational efficiency of its methods. The promoters of GreenAI suggested the use of Floating Point Operations (FLOPs) as a…

Accurately estimating workload runtime is a longstanding goal in computer systems, and plays a key role in efficient resource provisioning, latency minimization, and various other system management tasks. Runtime prediction is particularly…

Mixture-of-Experts (MoE) models, though highly effective for various machine learning tasks, face significant deployment challenges on memory-constrained devices. While GPUs offer fast inference, their limited memory compared to CPUs means…

Failure probability estimation problem is an crucial task in engineering. In this work we consider this problem in the situation that the underlying computer models are extremely expensive, which often arises in the practice, and in this…

As machine learning gets adopted into the industry quickly, trustworthiness is increasingly in focus. Yet, efficiency and sustainability of robust training pipelines still have to be established. In this work, we consider a simple pipeline…

We investigate the Optimal Obstacle Placement (OOP) problem under uncertainty, framed as the dual of the Optimal Traversal Path problem in the Stochastic Obstacle Scene paradigm. We consider both continuous domains, discretized for…

Edge Video Analytics (EVA) has gained significant attention as a major application of pervasive computing, enabling real-time visual processing. EVA pipelines, composed of deep neural networks (DNNs), typically demand efficient inference…

Deep learning is vulnerable to adversarial attacks, where carefully-crafted input perturbations could mislead a well-trained Deep Neural Network to produce incorrect results. Today's countermeasures to adversarial attacks either do not have…

Large language models (LLMs) power many state-of-the-art systems in natural language processing. However, these models are extremely computationally expensive, even at inference time, raising the natural question: when is the extra cost of…

Data science relies on pipelines that are organized in the form of interdependent computational steps. Each step consists of various candidate algorithms that maybe used for performing a particular function. Each algorithm consists of…

The most common approach to implementing data analysis pipelines involves obtaining point estimates from the upstream modules and then treating these as known quantities when working with the downstream ones. This approach is…

The training or fine-tuning of machine learning, vision, and language models is often implemented as a pipeline: a sequence of stages encompassing data preparation, model training and evaluation. In this paper, we exploit pipeline…

Making threaded programs safe and easy to reason about is one of the chief difficulties in modern programming. This work provides an efficient execution model for SCOOP, a concurrency approach that provides not only data race freedom but…

Simulations of exciton and charge hopping in amorphous organic materials involve numerous physical parameters. Each of these parameters must be computed from costly ab initio calculations before the simulation can commence, resulting in a…

Asynchronous pipeline model parallelism with a "1F1B" (one forward, one backward) schedule generates little bubble overhead and always provides quite a high throughput. However, the "1F1B" schedule inevitably leads to weight inconsistency…

When a large language model under reinforcement learning commits a wrong reasoning step early in a trajectory, standard algorithms force it to keep generating until the maximum horizon, spending compute on tokens that never receive positive…

Decision-based attacks construct adversarial examples against a machine learning (ML) model by making only hard-label queries. These attacks have mainly been applied directly to standalone neural networks. However, in practice, ML models…

Agentic workflows in large language model systems integrate retrieval, reasoning, and memory, but existing frameworks suffer from scalability and reproducibility limitations due to fragmented data orchestration, serialization overhead, and…

Multi-component natural language processing (NLP) pipelines are increasingly deployed for high-stakes decisions, yet no existing adversarial method can test their robustness under realistic conditions: binary-only feedback, no gradient…