Mark your calendar for Oct 13!
Join our live workshop at AI+X Summit 2023 and hear first hand from leading AI experts and industry veterans to explore the ins and outs of AI, its ethical implications and growing importance in our digital world. Among the array of topics, LatticeFlow’s innovative methodologies to make AI models safe and trustworthy will be a highlight. This is happening at a time when people are really asking for AI to be reliable, trustworthy, and fair. We will also dive deep into the EU AI Act and its implications on your business.
- October 13th, 2023
- 1:15 PM - 3:30 PM
- StageOne, Oerlikon
Agenda
Session 1: Trustworthy AI (60 min)
- 13:15-13:30 What you need to know about the EU AI Act: An insightful talk by Dr. Kevin Schawinski from Modulos.
- 13:30-13:45 Real-World AI Assessment for Migros Bank: Co-presented by Christian Rigassi from CRLT Services GmbH, mandated by Migros Bank, and Dr. Petar Tsankov from LatticeFlow.
- 13:45 - 14:15 Panel Discussion and Q&A: Get answers to your burning questions from our expert panel. Duration: 30 minutes.
- Moderator: Karin Schoefegger and (LatticeFLow).
Participants Kevin (Modulos), Christian (CRLT Services GmbH, mandated by Migros Bank), and Petar (LatticeFlow).
Session 2: LLM Security AI (60 min)
- 14:30-14:40 LLM Attacks Delve into the topic of "Mis-aligning Large Language Models" with our expert from ETH Zurich. Duration: 10 minutes.
- 14:40 - 14:50 LLM Safety "Open Problems and Emerging Solutions" - a deep dive by Mislav Balunovic from the SRI Lab. Duration: 10 minutes.
- 14:50 - 15:00 LMQL: Learn about "A language for efficient and reliable LLM inference" from Luca Beurer-Kellner of LMQL AI. Duration: 10 minutes.
- 15:00 - 15:10 Gandalf the Prompt Injection Game: Discover the "Learnings from 20 Million Prompts" with Václav Volhejn from Lakera. Duration: 10 minutes.
- 15:10 - 15:30 Panel Discussion and Q&A: Engage in an enlightening discussion and query session with our panel. Duration: 20 minutes.
- Moderator: Florian (ETH Zurich).
Participants Mislav & Luca (SRI Lab), Václav Volhejn (Lakera), and Pavol (LatticeFlow)
For more information:
- Event homepage:
- Workshop:
- Contact LatticeFlow:
Get to know the participants!
Petar Tsankov - CEO and Co-Founder of LatticeFlow.
Florian Tramèr - Assistant professor of computer science at ETH Zurich.
His research interests lie in Computer Security, Cryptography and Machine Learning security. In his current work, he studies the worst-case behavior of Deep Learning systems from an adversarial perspective, to understand and mitigate long-term threats on the safety and privacy of users.
Kevin Schawinski - Co-founder and CEO of Modulos
Kevin studied astrophysics at Oxford University, spent time as a researcher at Yale and NASA and worked as an assistant professor at ETH Zurich. Kevin is also the co-founder of the Galaxy Zoo citizen science project where he studied the growth of galaxies and supermassive black holes using space telescopes and pioneered approaches for using AI in astrophysics research.
Pavol Bielik - CTO and Co-Founder of LatticeFlow.
Pavol holds a PhD in machine learning and AI systems from ETH Zurich. He is also the honorary recipient of Meta’s Fellowship award, 2017 and the Romberg Grant, 2016. Pavol’s technical vision . As the CTO of LatticeFlow, Pavol is dedicated to pushing the boundaries of scaling the development of large machine learning models with a focus on enhancing quality and safety through automation.
Karin Schoefegger - Director of Product, LatticeFlow
Pavol holds a PhD in machine learning and AI systems from ETH Zurich. He is also the honorary recipient of Meta’s Fellowship award, 2017 and the Romberg Grant, 2016. Pavol’s technical vision . As the CTO of LatticeFlow, Pavol is dedicated to pushing the boundaries of scaling the development of large machine learning models with a focus on enhancing quality and safety through automation.
Mislav Balunovic - ETH Zurich
Mislav is a final year PhD student at ETH Zurich, supervised by Martin Vechev. The goal of his research is to develop robust, fair and private AI systems.
Václav Volhejn - Applied ML Scientist at Lakera
Václav’s current focus is on the research of prompt injection attacks and defenses. He is also one of the main creators of Gandalf, the world’s largest prompt injection game. Václav studied at Charles University in Prague and later relocated to Zürich to study CS at ETH on an ESOP scholarship. Václav earned a gold medal from the International Olympiad in Informatics in 2016.
Luca Beurer -Kellner - ETH Zurich
Lucca is a PhD student at the Department of Computer Science, ETH Zürich. His research revolves around machine learning and programming languages, with a focus on large language models. Luca is the main author of the LMQL programming language – an open source project that aims to improve the safety and reliability of language models, as they are deployed more widely.
