List your company

World Knowledge and Reading Comprehension on NQ, MMLU STEM, ARC, SciQ, LogiQA, BoolQ

29.81NQ AccuracyMistral-7B Base
12.608417.074221.5426.0058Feb 12, 2024

Evaluation results

MethodLinks
Mistral-7B BaseBackbone=Mistral-7B, Selection Method=None (Base Model), Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0229.8152.3984.6857.2597.430.2683.5854.3
DSIRBackbone=Mistral-7B, Selection Method=DSIR, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0229.2252.6284.7257.2597.330.2673.7653.54
Uniform (OpenWebMath)Backbone=Mistral-7B, Selection Method=Uniform Sampling, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0229.1752.1784.1856.6697.231.0383.8254.91
AutoDSBackbone=Mistral-7B, Selection Method=AutoDS, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0229.0652.384.1855.296.831.0383.1255.19
QuRatingBackbone=Mistral-7B, Selection Method=QuRating, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0228.8952.0185.4857.7697.331.1882.8154.03
Uniform (OpenWebMath)Backbone=LLaMA2-7B, Selection Method=Uniform Sampling, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0226.0740.0980.7750.0996.727.6578.4146.62
LLaMA2-7B BaseBackbone=LLaMA2-7B, Selection Method=None (Base Model), Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0226.0137.0880.7249.7496.826.5777.6844.99
QuRatingBackbone=LLaMA2-7B, Selection Method=QuRating, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0225.9637.8480.4350.696.827.6577.7144.97
AutoDSBackbone=LLaMA2-7B, Selection Method=AutoDS, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0225.8440.6680.0949.7496.727.9677.1947.07
DSIRBackbone=LLaMA2-7B, Selection Method=DSIR, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0225.7636.6380.9848.9896.526.7372.5444.27
Gemma-2B BaseBackbone=Gemma-2B, Selection Method=None (Base Model), Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0214.8836.677.6146.596.325.3569.5442.92
QuRatingBackbone=Gemma-2B, Selection Method=QuRating, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0214.3233.8177.9546.7696.224.4268.5341.67
Uniform (OpenWebMath)Backbone=Gemma-2B, Selection Method=Uniform Sampling, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0213.836.5477.445.3996.426.2768.3542.95
DSIRBackbone=Gemma-2B, Selection Method=DSIR, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0213.2734.4477.2745.8295.923.554.9239.71
AutoDSBackbone=Gemma-2B, Selection Method=AutoDS, Few-shot settings=NQ: 5, MMLU STEM: 5, ARC: 25, SciQ: 2, LogiQA: 2, BoolQ: 02024.0213.2736.0976.8146.0896.127.1971.2843.16