Would be interesting if this data could bring us ideas on information theory sort of how we have these nature parallels in algorithms like these: https://en.wikipedia.org/wiki/Ant_colony_optimization_algori...

Imagine Whales have better recall, pass@k metrics and deal with context windows differently, who knows!