Nikolas Gritsch

Björn Deiseroth, Max Meuer, Nikolas Gritsch et al.

Large Language Models (LLMs) have reshaped natural language processing with their impressive capabilities. However, their ever-increasing size has raised concerns about their effective deployment and the need for LLM compression. This study introduces the Divergent Token Metrics (DTMs), a novel approach to assessing compressed LLMs, addressing the limitations of traditional perplexity or accuracy measures that fail to accurately reflect text generation quality. DTMs measure token divergences that allow deeper insights into the subtleties of model compression, in particular, when evaluating components' impacts individually. Utilizing the First Divergent Token Metric (FDTM) in model sparsification reveals that 25% of all attention components can be pruned beyond 90% on the Llama-2 model family, still keeping SOTA performance. For quantization, FDTM suggests that more than 80% of parameters can be naively transformed to int8 without special outlier management. These evaluations indicate the necessity of choosing appropriate compressions for parameters individually -- and that FDTM can identify those -- while standard metrics result in deteriorated outcomes.

Nikolas Gritsch, Benjamin Tegeler, Faheem Hassan Zunjani

The world is under an ongoing pandemic, COVID-19, of a scale last seen a century ago. Contact tracing is one of the most critical and highly effective tools for containing and breaking the chain of infections especially in the case of infectious respiratory diseases like COVID-19. Thanks to the technological progress in our times, we now have digital mobile applications like the Corona-Warn-App for digital contact tracing. However, due to the invasive nature of contact tracing, it is very important to preserve the privacy of the users. Privacy preservation is important for increasing trust in the app and subsequently enabling its widespread usage in a privacy-valuing population. In this paper, we present a visual simulation of the working of the Corona-Warn-App to demonstrate how the privacy of its users is preserved, how they're notified of infectious contacts and how it helps in containing the spread of COVID-19.