Related papers: Adaptive Just-in-time Value Class Optimization for…

Intermediate features of a pre-trained model have been shown informative for making accurate predictions on downstream tasks, even if the model backbone is kept frozen. The key challenge is how to utilize these intermediate features given…

Popular approaches for minimizing loss in data-driven learning often involve an abstraction or an explicit retention of the history of gradients for efficient parameter updates. The aggregated history of gradients nudges the parameter…

Retrieval augmentation is a powerful but expensive method to make language models more knowledgeable about the world. Memory-based methods like LUMEN pre-compute token representations for retrieved passages to drastically speed up…

Value aggregation is a general framework for solving imitation learning problems. Based on the idea of data aggregation, it generates a policy sequence by iteratively interleaving policy optimization and evaluation in an online learning…

The computational cost of training multimodal large language models (MLLMs) grows rapidly with the number of processed tokens. Existing efficiency methods mainly target inference via token reduction or merging, offering limited benefits…

It is necessary to improve the performance of some special classes or to particularly protect them from attacks in adversarial learning. This paper proposes a framework combining cost-sensitive classification and adversarial learning…

Optimization problems with an auxiliary latent variable structure in addition to the main model parameters occur frequently in computer vision and machine learning. The additional latent variables make the underlying optimization task…

We introduce a novel latent vector space model that jointly learns the latent representations of words, e-commerce products and a mapping between the two without the need for explicit annotations. The power of the model lies in its ability…

Vision Transformers (ViTs) have shown success across a variety of tasks due to their ability to capture global image representations. Recent studies have identified the existence of high-norm tokens in ViTs, which can interfere with…

There has been recently a growing interest in studying adversarial examples on natural language models in the black-box setting. These methods attack natural language classifiers by perturbing certain important words until the classifier…

This paper introduces a type system for resource management in the context of nested virtualization. With nested virtualization, virtual machines compete with other processes for the resources of their host environment in order to provision…

The extensive memory footprint of language model (LM) fine-tuning poses a challenge for both researchers and practitioners. LMs use an embedding matrix to represent extensive vocabularies, forming a substantial proportion of the model…

A distinction is often drawn between a model's ability to predict a label for an evaluation sample that is directly memorised from highly similar training samples versus an ability to predict the label via some method of generalisation. In…

Incremental learning is useful if an AI agent needs to integrate data from a stream. The problem is non trivial if the agent runs on a limited computational budget and has a bounded memory of past data. In a deep learning approach, the…

In this paper, we contend that the objective of representation learning is to compress and transform the distribution of the data, say sets of tokens, towards a mixture of low-dimensional Gaussian distributions supported on incoherent…

Persistent Memory (PMEM), also known as Non-Volatile Memory (NVM), can deliver higher density and lower cost per bit when compared with DRAM. Its main drawback is that it is typically slower than DRAM. On the other hand, DRAM has…

Static program analysis by abstract interpretation is an efficient method to determine properties of embedded software. One example is value analysis, which determines the values stored in the processor registers. Its results are used as…

We introduce a video compression algorithm based on instance-adaptive learning. On each video sequence to be transmitted, we finetune a pretrained compression model. The optimal parameters are transmitted to the receiver along with the…

Training an effective video action recognition model poses significant computational challenges, particularly under limited resource budgets. Current methods primarily aim to either reduce model size or utilize pre-trained models, limiting…

Continual learning is crucial for applying machine learning in challenging, dynamic, and often resource-constrained environments. However, catastrophic forgetting - overwriting previously learned knowledge when new information is acquired -…