Table 2 Summary of the scenarios and subjects seen in the paper. Difficulty modalities as per Table 1. The last two columns show the mean normalised generality and its correlation with capability.
From: General intelligence disentangled via a generality metric for natural and artificial intelligence
Scenario | Subjects | Difficulty | Mean (\(\gamma\)) | Corr (\(\gamma\),\({\Psi }\)) |
|---|---|---|---|---|
Elithorn’s mazes | Humans | Intrinsic | 0.31 | 0.03 |
Letter series | Humans and machines | Intrinsic | 0.41 | 0.32 |
Object recognition | Humans, macaques and machines | Intrinsic | 0.54 | 0.92 |
Odour span task | Rats | Intrinsic | 0.28 | 0.16 |
Iris (KDM) | Machines | Intrinsic | 0.75 | − 0.32 |
Iris (\(TD_U\)) | Machines | Intrinsic | 0.84 | 0.11 |
Chess (Reykjavik) | Machines | Opp | 0.73 | 0.09 |
Chess (Leiden) | Machines | Opp | 0.41 | 0.32 |
ALE video games | Machines (+ human reference) | ARef | 0.88 | − 0.01 |
ALE video games | Machines and human | Rnk | 0.83 | − 0.17 |
GVGAI games (agg.) | Machines | Rnk | 0.78 | 0.02 |
Physical cognition | Orangutans | Rnk | 0.74 | 0.13 |
PCTB | Humans, chimpanzees and orangutans | DRef | 0.91 | 0.14 |
