Event announcement: Munich NLP Meetup

On Tuesday, February 10th, we will host the Munich NLP Meetup at our office in Munich. Join us for two talks:

First, our colleagues Henrik Klagges, Daniel Klingmann, and Fabian Klemm explain how TNG became one of the top 10 open-source model creators on OpenRouter. In their talk “AI Research @ TNG: How to process 20 billion tokens per day on OpenRouter”, they detail how TNG's Chimera LLMs were constructed from DeepSeek parent models.

Next, PhD student Ali Modarressi discusses how to measure memory failures of LLMs in his talk “Benchmarking Memory in LLMs: Retrieval, Long Context, and Multi-Turn Interaction”.

There will also be plenty of time for networking over pizza and drinks. More information and the free registration can be found here.