Pre-trained large language models, such as GPT\nobreakdash-2 and BERT, are often fine-tuned to achieve state-of-the-art performance on a downstream task. One natural example is the ``Smart Reply'' application where a pre-trained model is tuned to provide suggested responses for a given query message. Since the tuning data is often sensitive data such as emails or chat transcripts, it is important to understand and mitigate the risk that the model leaks its tuning data. We investigate potential information leakage vulnerabilities in a typical Smart Reply pipeline. We consider a realistic setting where the adversary can only interact with the underlying model through a front-end interface that constrains what types of queries can be sent to the model. Previous attacks do not work in these settings, but require the ability to send unconstrained queries directly to the model. Even when there are no constraints on the queries, previous attacks typically require thousands, or even millions, of queries to extract useful information, while our attacks can extract sensitive data in just a handful of queries. We introduce a new type of active extraction attack that exploits canonical patterns in text containing sensitive data. We show experimentally that it is possible for an adversary to extract sensitive user information present in the training data, even in realistic settings where all interactions with the model must go through a front-end that limits the types of queries. We explore potential mitigation strategies and demonstrate empirically how differential privacy appears to be a reasonably effective defense mechanism to such pattern extraction attacks.

Microsoft's Bing is enjoying the spotlight for the first time in a decade after it released a GPT-powered interface last month. But the tech giant has so far been cautious about the pace at which it is making the new Bing offering -- powered by OpenAI's GPT-4 tech -- available to users. But it appears, Bing is bringing those walls down. Microsoft, a major investor in OpenAI, appears to have lifted the waitlist from the new Bing, ostensibly allowing anyone to gain instant access to the new experience. Windows Central, which first spotted this change, said users don't have to wait to try out the new Bing anymore.

This article was published as a part of the Data Science Blogathon. In the first part of the series, we saw some most common techniques which we daily use while cleaning the data i.e. text cleaning in NLP. I would recommend if you haven't read it first read it, which will help you in text cleaning. The Link for the article is here. You can find the GitHub link here and start practicing and get your hand on the problem.

Let's face it; training a machine learning model is time-consuming. Even with the advancement in computing prowess over the past few years, training machine learning models takes a lot of time. Even the most trivial models have more than a million parameters. On a bigger scale, these models have over a billion parameters(GPT-3 has over 175 billion parameters!), and training these models takes days, if not weeks. As a Data Scientist, we would want to keep an eye on the model's metrics to know if the model performs as per expectations.

In this blog we shall learn how to create and share Microsoft Azure machine learning studio workspace using Microsoft Azure portal. For novice Cloud developers, and all other IT professionals associated with Cloud Big Data analytics especially with Microsoft Azure, this blog will help to get started with Microsoft Azure Machine learning. To enable computer understand from data and repetitive functional flow experiences along with making it to respond with no coding involved is, Machine learning.It helps to build powerful Artificial Intelligence (AI) applications which enables increase in speed and productivity helping organization to accomplish profitable targets. With continuing the same capabilities and feature, Machine Learning is now knowns as Machine Learning Studio. A powerful managed service enabling users to seamlessly build and share predictive analytics solutions.