About us
MEETUPS ARE TO BE HELD ON THE LAST WEDNESDAY OF EVERY MONTH
Turbine's AI meetups are meant for all researchers, engineers, scientists and students working on hard machine learning problems. It is created to dissect and understand new developments in ML together, and share our experience from real-life projects.
Presenters cover the latest impactful AI models, aiming to dive much deeper into each topic than what standard science communication formats allow. Thus, they'll expect you to have a working knowledge of machine learning.
In some sessions, we are going deep to understand recently published models and architectures - with working code whenever possible & intro to math background whenever needed. In others, presenters share their learnings working on models of real-life applications. Our goal is to give thorough knowledge to the audience, that you will use in model design on a daily basis.
Turbine is a computational biology company focusing on cancer, so expect lots of topics infused with biology. Yet, we are also a curious community, inviting you to join even if your personal interest centers around other domains. We also host completely biology-free events about computer vision, NLP and generic AI topics to cover the latest scientific advancements.
Select past presentations can be found here:
https://www.turbine.ai/ai-meetup-presentations
Upcoming events
1
The Platonic Representation Hypothesis
Turbine Kft., Szigony utca 26-32, Budapest, HUHumans can often move naturally between perception and language: we can look at something and describe it, or read something and form a mental picture. In machine learning, though, models for language and vision are usually trained separately, so it is easy to assume that they learn entirely different internal representations.
The Platonic Representation Hypothesis argues that this may not be true, and presents evidence that different models can converge toward similar underlying representations. In this meetup, we will walk through the paper, unpack its main ideas and results, and discuss the possible implications for how we think about intelligence, multimodal models, and the structure of learned representations.
Paper URL: https://arxiv.org/pdf/2405.07987
=== ENTRY DETAILS ===
- QR code with entry information will be available soon, in the "Photos" section of this event page.
- Gate closes at 18:15 - no late entries.49 attendees
Past events
25
