Fig. 3: Excerpt of state assignments in session 12 from mouse KS014, also shown in Fig. 2. | Nature Neuroscience

Fig. 3: Excerpt of state assignments in session 12 from mouse KS014, also shown in Fig. 2.

From: Infinite hidden Markov models can dissect the complexities of learning

Fig. 3

The left y axis serves as a scale for how connected a trial is to the other trials of that state (see Methods for details). The right y axis shows the contrast. The dot color indicates the animal’s response. One can see how the drastic and sudden change in the response patterns, rightwards (red) answers for leftwards (negative) contrasts, from trial ~330 to ~380 was detected by the model with a transition to state 7. The PMFs of states 4 and 6 looked similar but did, in fact, represent significantly higher error rates on the right and left sides, respectively. These mistakes are highlighted with arrows.

Back to article page