no code implementations • 17 May 2024 • Rya Sanovar, Srikant Bharadwaj, Renee St. Amant, Victor Rühle, Saravan Rajmohan
We identify that the associative property of online softmax can be treated as a reduction operation thus allowing us to parallelize the attention computation over these large context lengths.
1 code implementation • 19 Mar 2024 • Zhuoshi Pan, Qianhui Wu, Huiqiang Jiang, Menglin Xia, Xufang Luo, Jue Zhang, QIngwei Lin, Victor Rühle, Yuqing Yang, Chin-Yew Lin, H. Vicky Zhao, Lili Qiu, Dongmei Zhang
The challenge is that information entropy may be a suboptimal compression metric: (i) it only leverages unidirectional context and may fail to capture all essential information needed for prompt compression; (ii) it is not aligned with the prompt compression objective.
no code implementations • 28 Nov 2023 • Mohammad Mahdi Derakhshani, Menglin Xia, Harkirat Behl, Cees G. M. Snoek, Victor Rühle
We propose CompFuser, an image generation pipeline that enhances spatial comprehension and attribute assignment in text-to-image generative models.
no code implementations • 27 Nov 2023 • Lukas Wutschitz, Boris Köpf, Andrew Paverd, Saravan Rajmohan, Ahmed Salem, Shruti Tople, Santiago Zanella-Béguelin, Menglin Xia, Victor Rühle
In this paper, we take an information flow control perspective to describe machine learning systems, which allows us to leverage metadata such as access control policies and define clear-cut privacy and confidentiality guarantees with interpretable information flows.
1 code implementation • 10 Jun 2022 • Santiago Zanella-Béguelin, Lukas Wutschitz, Shruti Tople, Ahmed Salem, Victor Rühle, Andrew Paverd, Mohammad Naseri, Boris Köpf, Daniel Jones
Our Bayesian method exploits the hypothesis testing interpretation of differential privacy to obtain a posterior for $\varepsilon$ (not just a confidence interval) from the joint posterior of the false positive and false negative rates of membership inference attacks.
no code implementations • 12 Mar 2021 • FatemehSadat Mireshghallah, Huseyin A. Inan, Marcello Hasegawa, Victor Rühle, Taylor Berg-Kirkpatrick, Robert Sim
In this work, we introduce two privacy-preserving regularization methods for training language models that enable joint optimization of utility and privacy through (1) the use of a discriminator and (2) the inclusion of a triplet-loss term.
no code implementations • 14 Jan 2021 • Huseyin A. Inan, Osman Ramadan, Lukas Wutschitz, Daniel Jones, Victor Rühle, James Withers, Robert Sim
It has been demonstrated that strong performance of language models comes along with the ability to memorize rare training samples, which poses serious privacy threats in case the model is trained on confidential user content.
no code implementations • 17 Dec 2019 • Santiago Zanella-Béguelin, Lukas Wutschitz, Shruti Tople, Victor Rühle, Andrew Paverd, Olga Ohrimenko, Boris Köpf, Marc Brockschmidt
To continuously improve quality and reflect changes in data, machine learning applications have to regularly retrain and update their core models.