Dina BerenbaumRed Teaming AI: A Closer Look at PyRITDiscover how PyRIT can streamline your AI model evaluations. Dive deep into its modules and explore this open-source Microsoft library.9 min read·May 22, 2024----
Dina BerenbaumEnhancing JSON Output with Large Language Models: A Comprehensive GuideIn the realm of Large Language Models (LLMs), generating structured JSON outputs presents unique challenges. This guide explores various…9 min read·Jan 18, 2024--3--3
Dina BerenbaumLook ma, with hands! Using Handlebar.js templating to get exactly what you want from GPT4Yet another prompt engineering technique. But if you want to have your LLM output a very specific structure, you’ve come to the right…9 min read·Aug 21, 2023----
Dina BerenbaumCreating your own Q&A chatbot over your markdown documents using LangChainPractical step-by-step guide on how to use LangChain to create a personal or inner company chatbot. Examples using Azure and Weaviate.13 min read·Jun 15, 2023--2--2
Dina BerenbaumKeeping up with research: how to stay on top of things when time is pressing and the publications…The research in the area of Machine learning and Data science is exploding, both in academia and in the industry. In 2020, 54,160 papers…9 min read·Jan 19, 2022----
Dina BerenbaumMore with less: using pooling to detect coronavirus with fewer testsA week ago, as part of an academia-ministry of health brainstorming, I suggested to Prof. Roy Kishony, my former advisor, the idea of…7 min read·Mar 18, 2020--3--3
Dina BerenbauminTowards Data ScienceIs it easy to explain? Local explainabilityWhy should I trust your explanations? A more critical outlook at the most common explainability methods, their advantages but also…12 min read·Feb 2, 2020----
Dina BerenbauminTowards Data ScienceIs it easy to explain? Global explainability in Boosted TreesWhy should I trust your explanations? A more critical outlook at the most common explainability methods, their advantages but also…13 min read·Jan 23, 2020----
Dina BerenbauminTowards Data ScienceTake a deeper look at your Pytorch model with the new Tensorboard built-inDeep learning models are often complex, hard to understand and used as a “black box”. Making your model more accessible and visualize its…10 min read·Aug 25, 2019----