Calvin
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
CALVIN
100Success Rate
4
CALVIN ABC→D long-horizon
97Success Rate 1
4
CALVIN 10% of Env D
65.2No-RGB Success Rate
4
CALVIN Enriched instructions (D)
73.7Success Rate (Task 1)
4
CALVIN unseen lang
76.4Task Completion Rate (1 Task)
4
CALVIN 10% data
77.8Task 1 Completion Rate
4
CALVIN (test)
80.3Success Rate (1 task)
4
CALVIN Generalization 10% (transfer)
26Success Rate
3
CALVIN Generalization 10% (transfer)
33Success Rate
3
CALVIN
78.7Llama Q. Score
3
CALVIN
89.7Success Rate
3
CALVIN GPT-4 enriched Long-horizon (Split D)
73.6Success Rate (1/5)
3
CALVIN Severity 5
1.51Success Rate
2
CALVIN Severity 4
2.11Success Rate
2
CALVIN Severity 3
2.77Success Rate
2
CALVIN Clean
4.17Success Rate
2