List your company

ResearchFormerly WizWand

Reading Comprehension

Benchmarks

DatasetSOTA methodMetricTrendResultsLast updated

Llama-3.3-70B-Instruct

96Accuracy

13

May 4, 2026

66.51Accuracy

13

Feb 26, 2026

84.7BoolQ Score

12

Jul 2, 2026

Trinity Large (MoE)

84Accuracy

12

Feb 26, 2026

Belebele Korean (test)

90Accuracy

12

Feb 26, 2026

Belebele French

74.2Score

12

Feb 26, 2026

Reading Comprehension Benchmarks Aggregate

69.4Score

12

Feb 26, 2026

Children's Book Test (CBT) Named Entity (NE) (dev)

FG fine-grained gate

79.1Accuracy

12

Feb 26, 2026

Children's Book Test (CBT) Common Noun (CN) (dev)

FG fine-grained gate

75.3Accuracy

12

Feb 26, 2026

41.66Normalized Accuracy

11

Jun 29, 2026

72.1Normalized Accuracy

11

Jun 29, 2026

31.39Normalized Accuracy

11

Jun 29, 2026

59.19Exact Match (EM)

11

May 14, 2026

Belebele Indonesian 1.0 (test)

Trinity Large (MoE)

91Accuracy

11

Feb 26, 2026

Belebele Arabic 1.0 (test)

Trinity Large (MoE)

91Belebele Score

11

Feb 26, 2026

Belebele Russian (test)

Trinity Large (MoE)

92Accuracy

11

Feb 26, 2026

Belebele French (test)

Trinity Large (MoE)

92Accuracy

11

Feb 26, 2026

Belebele Japanese (test)

Trinity Large (MoE)

87Accuracy

11

Feb 26, 2026

Belebele Chinese (test)

Trinity Large (MoE)

92Belebele Accuracy

11

Feb 26, 2026

Belebele Portuguese (test)

Trinity Large (MoE)

91Accuracy

11

Feb 26, 2026

Belebele Vietnamese (test)

Trinity Large (MoE)

91Belebele Score

11

Feb 26, 2026

Instruction-Queryable LoRA

59.9Accuracy

11

May 12, 2026

DeepSeek Chat 7B

41.7EM

11

Feb 26, 2026

83.08Exact Match

11

Feb 26, 2026

DROP 1.0 (test)

Human Performance

92.38EM

11

Feb 26, 2026