Tasks
The benchmark comprises 23 tasks across various domains, assessing different skills of the model.
Type of task
All tasks
{{ taskName }}
Select domains
All domains
{{ domain.title }}
Expand the list of domains
Collapse the list of domains
Filters
Reset
Type of task
{{ taskName }}
Select domains
{{ domain.title }}
The name of the dataset
The ability of the model
Top Score | Human Baseline
Metric
{{ task.task_name }}
{{ task.scores.max }}
TS
{{ task.scores.human }}
HB
{{ metric.title }}
Description
Statistics
{{ key }}: {{ value }}
Output format
{{ format }}
Top score
Correctly
Good
Ethical
Virtue
{{ task.scores?.max['correct.virtue'] }}
{{ task.scores?.max['good.virtue'] }}
{{ task.scores?.max['ethical.virtue'] }}
Law
{{ task.scores?.max['correct.law'] }}
{{ task.scores?.max['good.law'] }}
{{ task.scores?.max['ethical.law'] }}
Moral
{{ task.scores?.max['correct.moral'] }}
{{ task.scores?.max['good.moral'] }}
{{ task.scores?.max['ethical.moral'] }}
Justice
{{ task.scores?.max['correct.justice'] }}
{{ task.scores?.max['good.justice'] }}
{{ task.scores?.max['ethical.justice'] }}
Utilitarianism
{{ task.scores?.max['correct.utilitarianism'] }}
{{ task.scores?.max['good.utilitarianism'] }}
{{ task.scores?.max['ethical.utilitarianism'] }}
Human Baseline
Correctly
Good
Ethical
Virtue
{{ task.scores?.human['correct.virtue'] }}
{{ task.scores?.human['good.virtue'] }}
{{ task.scores?.human['ethical.virtue'] }}
Law
{{ task.scores?.human['correct.law'] }}
{{ task.scores?.human['good.law'] }}
{{ task.scores?.human['ethical.law'] }}
Moral
{{ task.scores?.human['correct.moral'] }}
{{ task.scores?.human['good.moral'] }}
{{ task.scores?.human['ethical.moral'] }}
Justice
{{ task.scores?.human['correct.justice'] }}
{{ task.scores?.human['good.justice'] }}
{{ task.scores?.human['ethical.justice'] }}
Utilitarianism
{{ task.scores?.human['correct.utilitarianism'] }}
{{ task.scores?.human['good.utilitarianism'] }}
{{ task.scores?.human['ethical.utilitarianism'] }}
