1. Home
  2. Veranstaltungen
  3. Kolloquium
  4. On Trustworthiness of Large Language Models

On Trustworthiness of Large Language Models

Symbolbild Kolloquium

The past few years have witnessed significant leaps in capabilities of Large Language Models (LLMs). LLMs of today can perform a variety of tasks such as summarization, information retrieval and even mathematical reasoning with impressive accuracy. What is even more impressive is LLMs’ ability to follow natural language instructions without needing dedicated training datasets. However, issues like bias, hallucinations and lack of transparency pose a major impediment to wide adoption of these models. In this talk, I will review how we got from “traditional NLP” to today’s LLMs, and some of the reasons behind trustworthiness issues surrounding LLMs. I will then focus on a single issue — hallucinations in factual question answering — and show how artifacts associated with model generations can provide hints that the generation contains a hallucination.

Im Kolloquium stellen unsere Fellows, Arbeitsgemeinschatten und eingeladene Referent*innen regelmäßig ihre Forschungsprojekte vor. Gäste sind herzlich willkommen. Für eine bessere Planung senden Sie Ihre Anmeldung bitte an kolleg@cais-research.de

Eine Anfahrtsbeschreibung finden Sie unter Kontakt.

Weitere Termine

More events