Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

AgentBoard

Benchmarks

Task NameDataset NameSOTA ResultTrend
Agent task performanceAgentBoard ScienceWorld Naive Stream
Success Rate (1-step)66.7
4
Autonomous Agent Task SolvingAgentBoard Science World
Task Completion Rate58.9
3
Autonomous Agent Task SolvingAgentBoard WebShop
Task Completeness42.2
3
Autonomous Agent Task SolvingAgentBoard Tool-query Weather
Task Completeness75
3
Autonomous Agent Task SolvingAgentBoard Tool-query Academic
Task Completeness94.7
3
Autonomous Agent Task SolvingAgentBoard PDDL
Task Completion Rate76.7
3
Autonomous Agent Task SolvingAgentBoard Jericho
Task Completeness50
3
Autonomous Agent Task SolvingAgentBoard Babyai
Task Completeness100
3
Autonomous Agent Task SolvingAgentBoard AlfWorld
Task Completion Rate63.4
3
Showing 9 of 9 rows