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

WinoGrande

Benchmarks

Task NameDataset NameSOTA ResultTrend
Commonsense ReasoningWinoGrande
Accuracy7,364
1,085
Commonsense ReasoningWinogrande
Accuracy85.3
372
Common sense reasoningWinogrande
Accuracy91.3
189
ReasoningWinoGrande (WG)
Accuracy85.2
135
Question AnsweringWinogrande (WG)
Accuracy72.77
124
Commonsense ReasoningWinoGrande (val)
Accuracy73.88
87
Commonsense ReasoningWinogrande
Accuracy76.09
78
Commonsense ReasoningWinogrande
Accuracy69.77
68
Commonsense ReasoningWinoGrande 5-shot
Accuracy92.66
64
Zero-shot ReasoningWinoGrande
Accuracy70
54
Commonsense ReasoningWinogrande
Accuracy (0-shot)73.7
42
Pronoun ResolutionWinoGrande
Accuracy89.4
41
Coreference ResolutionWinogrande
Accuracy73.2
40
Commonsense ReasoningWinoGrande standard (test)
Accuracy80.2
35
Identification of inactive attention headsWinoGrande
Percentage of Zeroed Heads20.91
30
Zero-shot AccuracyWinoGrande
Zero-shot Accuracy77.3
30
Commonsense Question AnsweringWinoGrande (WG) (val)
Accuracy78.3
21
LLM Performance EstimationWinogrande (test)
MAE1.027
20
Commonsense ReasoningWinogrande
Accuracy0.8624
19
Commonsense ReasoningWinogrande
Accuracy (Pre-Attack)73.2
19
Multiple-choice commonsense reasoningWinogrande
Winogrande Accuracy74
18
Commonsense ReasoningWinogrande
HS (Head-to-Head Score)47.75
17
Zero-shot PredictionWinogrande
Accuracy69.06
17
Commonsense ReasoningWinogrande
Accuracy88.8
16
Common Sense ReasoningWinoGrande (dev)
Accuracy79.2
16
Showing 25 of 67 rows