ML Pub Club #16: Reasoning in LLMs and their limitations

​How is reasoning in LLMs achieved and to they reason at all? What are the limitations of the current models? Scaling laws, impact of model size on reasoning and some intuitions behind it.

​How to perform scientific research and controlled experiments with LLMs? RL with LLMs is all the rage now - what is that about? What are the current research directions and what does the future bring? 🤔

​To answer all these question, we will be welcoming Marko Velić Engineering Manager at Google, to our upcoming ML Pub Club #16.

​Marko previously worked at Photomath, Meta (Facebook), Styria, LEGO (as consultant), University of Zagreb and has entrepreneurship experience. He won many awards including the ones from Microsoft and NVIDIA, and is also a lecturer at the Algebra University. He is interested in ML research and practical applications in real life scenarios.

​[This event will be held in English.]