TEPCO continues to release very limited data sets, which are very difficult to interpret, but they appear to show that Iodine-131 has been produced by recent criticality events. TEPCO has provided data on contamination by Iodine-131 with a half life of 8 days, Cesium-134 with a half life of 2 years and Cesium-137 with a half life of 30 years. Because we don't know the sources, sinks, the amounts of dilution by water and the amounts of concentration by evaporation, the best data for interpretation involve different isotopes, with different half lives of the same element. The ratio of a shorter lived isotope of an element compared to a longer lived isotope of that element should decline exponentially over time. With isotopes of different elements, the ratios will decline systematically with time if chemical and dilution processes don't change over time. That's just what we see with the new data on the radioactivity of sea water at Iwasawa shore. The amount of short lived Iodine-131 found in sea water has declined much faster than the amount of longer lived Cesium-134 and Cesium-137.
However, inside of the silt fence near Reactor Unit 2 both the absolute levels of Iodine-131 and the ratios of iodine to cesium increase unexpectedly at April 15 and April 25. Less dilution with uncontaminated water could cause the absolute level of Iodine-131 to increase, but it would not likely increase the isotope ratio of iodine to cesium. I can think of actions by TEPCO that could produce that unexpected result, but I think they would have reported such actions because it would have made their data look less concerning. The simplest explanation for these data is that reactor 2 went critical on April 15 and April 25.
It's possible the contaminated water migrated from reactor 3 or reactor 1 or that TEPCO used an adsorbent for clean up that took up cesium but not iodine. If we had measurements of other Iodine isotopes we could be more sure of the cause of the 2 unexpected spikes in short lived Iodine-131.
Moreover, the subdrain of Unit 3 shows an unexpected increase in both the absolute level and the isotope ratios of Iodine-131 that began about April 20. The increase of I-131 by about a factor of 10, while cesium levels spiked, then stabilized, is hard to explain if reactor 3 (or another reactor) didn't go critical on about April 20.
Several other recently released figures show similar unexplained increases in I-131. The limited data TEPCO has released support the hypothesis that reactor units 2 and 3 are not under control. There is evidence for 3 possible criticality events in April.