Reading Comprehension on BIG-bench

Metric: Accuracy (higher is better)

LeaderboardDataset

Loading chart...

Results

Sort:

#	Model↕	Accuracy ▼	Extra Data	Paper	Date↕	Code
1	Chinchilla-70B (few-shot, k=5)	94	No	Training Compute-Optimal Large Language Models	2022-03-29	Code
2	Chinchilla-70B (few-shot, k=5)	92.8	No	Training Compute-Optimal Large Language Models	2022-03-29	Code
3	Gopher-280B (few-shot, k=5)	88.7	No	Scaling Language Models: Methods, Analysis & Ins...	2021-12-08	Code
4	Chinchilla-70B (few-shot, k=5)	82.4	No	Training Compute-Optimal Large Language Models	2022-03-29	Code
5	Gopher-280B (few-shot, k=5)	81.8	No	Scaling Language Models: Methods, Analysis & Ins...	2021-12-08	Code
6	Chinchilla-70B (zero-shot)	77.4	No	Training Compute-Optimal Large Language Models	2022-03-29	Code
7	Gopher-280B (few-shot, k=5)	75.1	No	Scaling Language Models: Methods, Analysis & Ins...	2021-12-08	Code
8	Gopher-280B (zero-shot)	74.5	No	Scaling Language Models: Methods, Analysis & Ins...	2021-12-08	Code
9	Chinchilla-70B (few-shot, k=5)	69	No	Training Compute-Optimal Large Language Models	2022-03-29	Code
10	Gopher-280B (few-shot, k=5)	64.1	No	Scaling Language Models: Methods, Analysis & Ins...	2021-12-08	Code
11	Chinchilla-70B (few-shot, k=5)	63.3	No	Training Compute-Optimal Large Language Models	2022-03-29	Code
12	Gopher-280B (few-shot, k=5)	57.6	No	Scaling Language Models: Methods, Analysis & Ins...	2021-12-08	Code
13	Chinchilla-70B (few-shot, k=5)	54.5	No	Training Compute-Optimal Large Language Models	2022-03-29	Code
14	Chinchilla-70B (few-shot, k=5)	53.1	No	Training Compute-Optimal Large Language Models	2022-03-29	Code
15	Gopher-280B (few-shot, k=5)	52.7	No	Scaling Language Models: Methods, Analysis & Ins...	2021-12-08	Code
16	Gopher-280B (few-shot, k=5)	50.7	No	Scaling Language Models: Methods, Analysis & Ins...	2021-12-08	Code
17	Chinchilla-70B (few-shot, k=5)	49.4	No	Training Compute-Optimal Large Language Models	2022-03-29	Code
18	Gopher-280B (few-shot, k=5)	36.4	No	Scaling Language Models: Methods, Analysis & Ins...	2021-12-08	Code
19	Gopher-280B (few-shot, k=5)	27.3	No	Scaling Language Models: Methods, Analysis & Ins...	2021-12-08	Code

#1Chinchilla-70B (few-shot, k=5)SOTA
94
Accuracy · 2022-03-29
Training Compute-Optimal Large Language Models Code
#2Chinchilla-70B (few-shot, k=5)
92.8
Accuracy · 2022-03-29
Training Compute-Optimal Large Language Models Code
#3Gopher-280B (few-shot, k=5)SOTA
88.7
Accuracy · 2021-12-08
Scaling Language Models: Methods, Analysis & Insights from Training Gopher Code
#4Chinchilla-70B (few-shot, k=5)
82.4
Accuracy · 2022-03-29
Training Compute-Optimal Large Language Models Code
#5Gopher-280B (few-shot, k=5)
81.8
Accuracy · 2021-12-08
Scaling Language Models: Methods, Analysis & Insights from Training Gopher Code
#6Chinchilla-70B (zero-shot)
77.4
Accuracy · 2022-03-29
Training Compute-Optimal Large Language Models Code
#7Gopher-280B (few-shot, k=5)
75.1
Accuracy · 2021-12-08
Scaling Language Models: Methods, Analysis & Insights from Training Gopher Code
#8Gopher-280B (zero-shot)
74.5
Accuracy · 2021-12-08
Scaling Language Models: Methods, Analysis & Insights from Training Gopher Code
#9Chinchilla-70B (few-shot, k=5)
69
Accuracy · 2022-03-29
Training Compute-Optimal Large Language Models Code
#10Gopher-280B (few-shot, k=5)
64.1
Accuracy · 2021-12-08
Scaling Language Models: Methods, Analysis & Insights from Training Gopher Code
#11Chinchilla-70B (few-shot, k=5)
63.3
Accuracy · 2022-03-29
Training Compute-Optimal Large Language Models Code
#12Gopher-280B (few-shot, k=5)
57.6
Accuracy · 2021-12-08
Scaling Language Models: Methods, Analysis & Insights from Training Gopher Code
#13Chinchilla-70B (few-shot, k=5)
54.5
Accuracy · 2022-03-29
Training Compute-Optimal Large Language Models Code
#14Chinchilla-70B (few-shot, k=5)
53.1
Accuracy · 2022-03-29
Training Compute-Optimal Large Language Models Code
#15Gopher-280B (few-shot, k=5)
52.7
Accuracy · 2021-12-08
Scaling Language Models: Methods, Analysis & Insights from Training Gopher Code
#16Gopher-280B (few-shot, k=5)
50.7
Accuracy · 2021-12-08
Scaling Language Models: Methods, Analysis & Insights from Training Gopher Code
#17Chinchilla-70B (few-shot, k=5)
49.4
Accuracy · 2022-03-29
Training Compute-Optimal Large Language Models Code
#18Gopher-280B (few-shot, k=5)
36.4
Accuracy · 2021-12-08
Scaling Language Models: Methods, Analysis & Insights from Training Gopher Code
#19Gopher-280B (few-shot, k=5)
27.3
Accuracy · 2021-12-08
Scaling Language Models: Methods, Analysis & Insights from Training Gopher Code