Zobrazeno 1 - 10
of 2 078
pro vyhledávání: '"Mullins, Robert A."'
Autor:
Khachaturov, David, Mullins, Robert
Quantifying robustness in a single measure for the purposes of model selection, development of adversarial training methods, and anticipating trends has so far been elusive. The simplest metric to consider is the number of trainable parameters in a m
Externí odkaz:
http://arxiv.org/abs/2410.18556
Autor:
Zhang, Zixi, Zhang, Cheng, Gao, Xitong, Mullins, Robert D., Constantinides, George A., Zhao, Yiren
Low-rank Adaption (LoRA) has been the de-facto parameter-efficient fine-tuning technique for large language models. We present HeteroLoRA, a light-weight search algorithm that leverages zero-cost proxies to allocate the limited LoRA trainable paramet
Externí odkaz:
http://arxiv.org/abs/2406.14956
Autor:
Chen, Yuang, Zhang, Cheng, Gao, Xitong, Mullins, Robert D., Constantinides, George A., Zhao, Yiren
Grouped-query attention (GQA) has been widely adopted in LLMs to mitigate the complexity of multi-head attention (MHA). To transform an MHA to a GQA, neighbour queries in MHA are evenly split into groups where each group shares the value and key laye
Externí odkaz:
http://arxiv.org/abs/2406.14963
Deep neural networks, costly to train and rich in intellectual property value, are increasingly threatened by model extraction attacks that compromise their confidentiality. Previous attacks have succeeded in reverse-engineering model parameters up t
Externí odkaz:
http://arxiv.org/abs/2406.10011
Feedback data plays an important role in fine-tuning and evaluating state-of-the-art AI models. Often pairwise text preferences are used: given two texts, human (or AI) annotators select the "better" one. Such feedback data is widely used to align mo
Externí odkaz:
http://arxiv.org/abs/2406.06560
Autor:
Clifford, Eleanor, Saravanan, Adhithya, Langford, Harry, Zhang, Cheng, Zhao, Yiren, Mullins, Robert, Shumailov, Ilia, Hayes, Jamie
Modern Machine Learning models are expensive IP and business competitiveness often depends on keeping this IP confidential. This in turn restricts how these models are deployed -- for example it is unclear how to deploy a model on-device without inev
Externí odkaz:
http://arxiv.org/abs/2405.20990
While previous research backdoored neural networks by changing their parameters, recent work uncovered a more insidious threat: backdoors embedded within the definition of the network's architecture. This involves injecting common architectural compo
Externí odkaz:
http://arxiv.org/abs/2402.06957
Test stimuli generation has been a crucial but labor-intensive task in hardware design verification. In this paper, we revolutionize this process by harnessing the power of large language models (LLMs) and present a novel benchmarking framework, LLM4
Externí odkaz:
http://arxiv.org/abs/2310.04535
Autor:
Khachaturov, David, Gao, Yue, Shumailov, Ilia, Mullins, Robert, Anderson, Ross, Fawaz, Kassem
Visual adversarial examples have so far been restricted to pixel-level image manipulations in the digital world, or have required sophisticated equipment such as 2D or 3D printers to be produced in the physical real world. We present the first ever m
Externí odkaz:
http://arxiv.org/abs/2310.00438
Current literature demonstrates that Large Language Models (LLMs) are great few-shot learners, and prompting significantly increases their performance on a range of downstream tasks in a few-shot learning setting. An attempt to automate human-led pro
Externí odkaz:
http://arxiv.org/abs/2304.03609