数据集:
lighteval/mmlu
任务:
问答子任务:
multiple-choice-qa语言:
en计算机处理:
monolingual大小:
10K<n<100K语言创建人:
expert-generated批注创建人:
no-annotation源数据集:
original许可:
mitMeasuring Massive Multitask Language Understanding by Dan Hendrycks , Collin Burns , Steven Basart , Andy Zou, Mantas Mazeika, Dawn Song , and Jacob Steinhardt (ICLR 2021).
This is a massive multitask test consisting of multiple-choice questions from various branches of knowledge. The test spans subjects in the humanities, social sciences, hard sciences, and other areas that are important for some people to learn. This covers 57 tasks including elementary mathematics, US history, computer science, law, and more. To attain high accuracy on this test, models must possess extensive world knowledge and problem solving ability.
A complete list of tasks: ['abstract_algebra', 'anatomy', 'astronomy', 'business_ethics', 'clinical_knowledge', 'college_biology', 'college_chemistry', 'college_computer_science', 'college_mathematics', 'college_medicine', 'college_physics', 'computer_security', 'conceptual_physics', 'econometrics', 'electrical_engineering', 'elementary_mathematics', 'formal_logic', 'global_facts', 'high_school_biology', 'high_school_chemistry', 'high_school_computer_science', 'high_school_european_history', 'high_school_geography', 'high_school_government_and_politics', 'high_school_macroeconomics', 'high_school_mathematics', 'high_school_microeconomics', 'high_school_physics', 'high_school_psychology', 'high_school_statistics', 'high_school_us_history', 'high_school_world_history', 'human_aging', 'human_sexuality', 'international_law', 'jurisprudence', 'logical_fallacies', 'machine_learning', 'management', 'marketing', 'medical_genetics', 'miscellaneous', 'moral_disputes', 'moral_scenarios', 'nutrition', 'philosophy', 'prehistory', 'professional_accounting', 'professional_law', 'professional_medicine', 'professional_psychology', 'public_relations', 'security_studies', 'sociology', 'us_foreign_policy', 'virology', 'world_religions']
Model | Authors | Humanities | Social Science | STEM | Other | Average |
---|---|---|---|---|---|---|
UnifiedQA | Khashabi et al., 2020 | 45.6 | 56.6 | 40.2 | 54.6 | 48.9 |
GPT-3 (few-shot) | Brown et al., 2020 | 40.8 | 50.4 | 36.7 | 48.8 | 43.9 |
GPT-2 | Radford et al., 2019 | 32.8 | 33.3 | 30.2 | 33.1 | 32.4 |
Random Baseline | N/A | 25.0 | 25.0 | 25.0 | 25.0 | 25.0 |
English
An example from anatomy subtask looks as follows:
{ "question": "What is the embryological origin of the hyoid bone?", "choices": ["The first pharyngeal arch", "The first and second pharyngeal arches", "The second pharyngeal arch", "The second and third pharyngeal arches"], "answer": "D" }
auxiliary_train | dev | val | test | |
---|---|---|---|---|
TOTAL | 99842 | 285 | 1531 | 14042 |
Transformer models have driven this recent progress by pretraining on massive text corpora, including all of Wikipedia, thousands of books, and numerous websites. These models consequently see extensive information about specialized topics, most of which is not assessed by existing NLP benchmarks. To bridge the gap between the wide-ranging knowledge that models see during pretraining and the existing measures of success, we introduce a new benchmark for assessing models across a diverse set of subjects that humans learn.
[More Information Needed]
Who are the source language producers?[More Information Needed]
[More Information Needed]
Who are the annotators?[More Information Needed]
[More Information Needed]
[More Information Needed]
[More Information Needed]
[More Information Needed]
[More Information Needed]
If you find this useful in your research, please consider citing the test and also the ETHICS dataset it draws from:
@article{hendryckstest2021, title={Measuring Massive Multitask Language Understanding}, author={Dan Hendrycks and Collin Burns and Steven Basart and Andy Zou and Mantas Mazeika and Dawn Song and Jacob Steinhardt}, journal={Proceedings of the International Conference on Learning Representations (ICLR)}, year={2021} } @article{hendrycks2021ethics, title={Aligning AI With Shared Human Values}, author={Dan Hendrycks and Collin Burns and Steven Basart and Andrew Critch and Jerry Li and Dawn Song and Jacob Steinhardt}, journal={Proceedings of the International Conference on Learning Representations (ICLR)}, year={2021} }
Thanks to @andyzoujm for adding this dataset.