History bias and its perturbation of the stimulus representation in the macaque prefrontal cortex.
Benozzo, D.; Ferrucci, L.; Ceccarelli, F.; Genovesio, A.
Show abstract
Multiple history biases affect our representation of magnitudes, such as time, distance, and size. It is not clear whether the previous stimuli interfere with the discrimination process from the moment of stimulus presentation, during working memory retention, or even later during the decision-making phase. We used a spatial discrimination task involving two stimuli of different magnitudes, presented sequentially at various distances from the center. The monkeys task was to select the farthest of them. We showed that the previous stimulus magnitude generated a contraction bias effect, but only when its stimulus features differed from those of the current stimulus. In this case, at the neural level we also observed that the decoding of the stimulus magnitude achieved the highest accuracy when it matched the magnitude of the preceding stimulus for which the decoder was trained. This indicates that past stimuli can affect magnitude processing already during the stimulus presentation, even before the decision process. Interestingly, this effect manifested when the trace of the previous stimulus magnitude reactivated in the second part of the stimulus presentation after an "activity-silent" period.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- The Spatial Frequency Representation Predicts Category Coding in the Inferior Temporal Cortex 96%
- Location- and feature-based selection histories make independent, qualitatively distinct contributions to urgent visuomotor performance 95%
- A Bayesian and efficient observer model explains concurrent attractive and repulsive history biases in visual perception 95%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.