Fig. 1: List functions vary widely in difficulty and algorithmic content. | Nature Communications

Fig. 1: List functions vary widely in difficulty and algorithmic content.

From: Symbolic metaprogram search improves learning efficiency and explains rule learning in humans

Fig. 1

Six list functions with an English description, human mean accuracy (n = 389 people) in parentheses, and input → output examples. Plot shows empirical distribution over accuracy per function (100 functions) for humans (darker = more mass); dots show mean accuracy with example functions marked in blue.

Back to article page