Overcooked
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Overcooked
607.76IQM Return
5
Overcooked 1.0 (Extended Evaluation 10 rounds x 25 episodes)
93.5Success Rate
5
Overcooked Short evaluation (10 rounds x 5 episodes) 1.0
90Success Rate
5
Overcooked Coordination Ring layout
96Classification Accuracy
5
Overcooked Asymmetric Advantage layout
77Classification Accuracy
5
Overcooked Cramped Room layout
96Accuracy
5
Overcooked Forced Coord.
43.8Total Mean Reward
4
Overcooked Counter Circuit
44.38Total Mean Reward
4
Overcooked Coord. Ring
105.63Mean Reward
4
Overcooked Asym. Adv.
112.5Total Mean Reward
4
Overcooked Forced Coord.
35Mean Reward
4
Overcooked Counter Circuit
63Mean Reward
4
Overcooked Coord. Ring
60Mean Reward
4
Overcooked Asym. Adv.
198.18Mean Reward
4
Overcooked Cramped Room
118.18Mean Reward
4
Overcooked Coord. Ring layout (Human Study)
63.5Return
4
Overcooked (OC)
152.31Base Score
4
Overcooked Task Generalization (held-out)
85.96Coaching Score (LLM-as-Judge)
4
Overcooked Defects (held-out)
77.8Coaching (LLM-as-judge)
4
Overcooked Cramped Room
91.33Reward
4
Overcooked Coordination Ring
8.4Reward
4
Overcooked Asymmetric Advantages
72Reward
4
Overcooked Asym.-2
97Floor Score
3
Overcooked (Cramped-2)
99Floor Score
3
Overcooked Asym.-1
99Floor Score
3