Fig. 6.

Inspection of activations for the [i-i] pair for the non-identical cues. Activation (top row) and average cue weight (bottom row) for only those cues which are not identical to the target. See Fig. 4 for how to read the GAM plot. Activation (top row) seems to go in the opposite to expected direction. However, this is because there are a greater number of non-identical cues for continuum steps further from the target. The average cue weight (bottom row) shows a similar pattern to the main results in Fig. 4: there is a gradient change in activation over the continuum. This result shows that the change in activation across the continuum is not limited to identical cues, but also occurs for spectral components that differ between the test stimulus and target stimulus.