1 code implementation • 22 Apr 2023 • Shima Rahimi Moghaddam, Christopher J. Honey
GPT-4 performed best in zero-shot settings, reaching nearly 80% ToM accuracy, but still fell short of the 87% human accuracy on the test set.
no code implementations • 12 Dec 2020 • Shima Rahimi Moghaddam, Fanjun Bu, Christopher J. Honey
How does the temporal "smoothness" of internal states affect the efficiency of learning when the training data are also temporally smooth?