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

Mean Reward on ALFWorld

0.767Mean Reward

NO-COMP

Updated 7d ago

Evaluation Results

Method	Links
NO-COMP 2026.05		0.767	-
AGORA 2026.05		0.767	100
Trunc 2026.05		0.767	100
NO-COMP 2026.05		0.767	-
AGORA 2026.05		0.633	83
Trunc 2026.05		0.633	83
Floor-K2 2026.05		0.6	78
Floor-K2 2026.05		0.5	65
LongLingua 2026.05		0.467	100
LongLingua 2026.05		0.467	93
HiAgent 2026.05		0.4	86
ACON 2026.05		0.4	86
HiAgent 2026.05		0.4	80
AgentDiet 2026.05		0.333	71
AGORA 2026.05		0.333	111
ObsMask 2026.05		0.3	64
NO-COMP 2026.05		0.3	-
Trunc 2026.05		0.267	89
ObsMask 2026.05		0.267	53
AgentDiet 2026.05		0.267	53
Floor-K2 2026.05		0.233	78
ACON 2026.05		0.233	47
AgentDiet 2026.05		0.167	71
ACON 2026.05		0.167	71
ObsMask 2026.05		0.1	43
LongLingua 2026.05		0.1	43
HiAgent 2026.05		0.1	43
SelCtx 2026.05		0.033	7
SelCtx 2026.05		0	0
SelCtx 2026.05		0	0