Fig. 2: Interpretation of Individual GLM Results Using an LLM (GPT-4o).
From: Personalised modelling of routine variability and affective states
We used an LLM (OpenAI’s GPT-4o (temperature=0, max token=4096)) to translate an individual’s anxiety GLM results into user-accessible language along with the routine definition file. We presented the actual output from the LLM.
