ForecastBench-Sim

Advanced LLMs show inverse scaling in forecasting, study finds

A new research paper released on arXiv suggests that more capable large language models may perform worse on forecasting tasks involving financial or epidemiological data. The study found that these advanced models tend…