These graphs show the aggregated classification over three different runs, where for every category (FN, TP, TN, FP) the minimal number over the different runs is taken. For example, if 500 traces are TP in the first run, 490 in the second, and 495 in the third, then the value 490 is used as aggregated value. As a result, these graphs can show discrepancies between the classifications in different runs.
For the Inductive Miner (OR), there are discrepancies (note that some bars do not reach 1000). There are many traces that cannot be replayed in the discovered nets within 10 seconds. As a result, the discrepancy can either be caused by different replays, or by different discovered nets. For some logs where a discrepancy occurs, the Inductive Miner (OR) indeed discovered different nets from this log over the three runs.