20 choice trials completed, and then averaging those values across 3 sessions within a programmed delay. Sessions were limited to 1 hour or 60 choice trials, whichever came first. All daily data from mice that completed fewer than 20 trials on a day were excluded. If a mouse did not complete ≥20 trials on 2 of 3 sessions at a delay, that mouse was excluded from the experiment for nonperformance.
