Elina Lesyk
Elina Lesyk is an Associate Solutions Architect at AWS located in Munich. Coming to an IT area as to the most efficient problem solver and decision driver, she is exploring the world of the cloud, AI and sustainability with Financial Services Enterprise customers. Excited to talk about Generative AI, AI for Good, ESG and the cloud! In her free time, Elina likes learning guitar theory in Spanish to cross-learn and going for some planned and unplanned half-marathons.
Make Large Language Models Behave: 7 Ideas for Builders
Have we opened Pandora's black box with Transformers? How much can be leaked through smart prompts? Is it possible to make the evil behave? Let's explore several prompts that you'd want to put into an LLM at hand and see how it reacts. After wearing the malicious hacker hat, we'll look into soft and hard defensive measures against prompt hacking. Of course with the hands-on on how to build into them your Transformer-based solution. I want your LLMs to generate polite, secure and inclusive responses!