Can You Answer This? – Exploring Zero-Shot QA Generalization Capabilities in Large Language Models (Student Abstract)

The buzz around Transformer-based language models (TLM) such as BERT, RoBERTa, etc. is well-founded owing to their impressive results on an array of tasks. However, when applied areas needing specialized knowledge (closed-domain), medical, finance, performance takes drastic hits, sometimes more than older recurrent/convolutional counterparts. In this paper, we explore zero-shot capabilities large LMs for extractive QA. Our objective examine change in the face domain drift i.e. target data vastly different semantic and statistical properties from source attempt explain subsequent behavior. To end, present two studies paper while planning further experiments later down road. findings indicate flaws current generation TLM limiting closed-domain