Fig. 2: diHMM fit to mouse KS014.
From: Infinite hidden Markov models can dissect the complexities of learning
The topmost row shows the overall performance during the session, as proportion correct, and the current stage of learning as the background color (we elaborate on learning stages later in the text). Vertical lines with shaded circles at the top indicate the sessions during which new contrasts were introduced. The remaining rows show the prevalent behavioral states (label to the right) ordered by appearance, indicating which proportion of trials they explained during each session. To the far right of every state, we show its PMFs across time, ignoring the contribution from the history of previous choices. The saturation of the colors of the states indicates successive appearance and matches the PMF plots.
