Join Anders Larson and Shea Parkes as they delve into the technical aspects of large language models (LLMs); this episode focuses specifically on the process of tokenization. They discuss how understanding the workings of these models can aid in interpreting their outputs and integrating them into actuarial work. The conversation covers the evolution of natural language processing, the significance of tokens in LLMs for bridging text to numbers, and the stochastic process of generating new tokens.
SOA Podcasts - Society of Actuaries