Value cue | Delay | Saccade | Reward | Total | |
---|---|---|---|---|---|
Task-related responses | 149 | 97 | 87 | 97 | 430 |
A: 101 | A: 65 | A: 50 | A: 60 | A: 276 | |
B: 48 | B: 32 | B: 37 | B: 37 | B: 154 | |
Value-coding responses | 63 (42.3%) | 40 (41.2%) | 19 (21.8%) | 27 (27.8%) | 149 (34.7%) |
A: 40 (39.6%) | A: 27 (41.5%) | A: 10 (20.0%) | A: 11 (18.3%) | A: 88 (31.9%) | |
B: 23 (47.9%) | B: 13 (40.6%) | B: 9 (24.3%) | B: 16 (43.2%) | B: 61 (39.6%) | |
Adaptive responses | 22 (34.9%) | 12 (30%) | 1 (5.3%) | 3 (11.1%) | 38 (25.5%) |
A: 16 (40%) | A: 11 (40.7%) | A: 1 (10.0%) | A: 2 (18.2%) | A: 30 (34.1%) | |
B: 6 (26.1%) | B: 1 (7.7%) | B: 0 (0%) | B: 1 (6.3%) | B: 8 (13.1%) |
Top, Task-related responses. Responses significantly different from baseline (intertrial interval) activity were counted as task-related responses in each task period (p < 0.01 post hoc Scheffé test). Middle, Value-coding responses. Responses significantly correlated with predicted juice volume were counted as value-coding responses in each task period (p < 0.05, Spearman's rank correlation, corrected for multiple comparisons). Numbers in parentheses indicate proportions of value-coding responses in task-related responses during each task period. Bottom, Adaptive responses. Neuronal adaptation was examined by comparing the regression slopes between narrow and wide reward distributions using regression models (Eqs. 1, 2). Numbers in parentheses indicate adaptive responses as percentage of value-coding responses during each task period. A and B refer to the two animals used.