Leo Gao
First author of The Pile and first-listed author of EleutherAI's LM Evaluation Harness
Leo Gao is first author of The Pile, an 825 GiB English text corpus for training language models that was published with an EleutherAI affiliation in 2020. Gao is the first-listed author in the citation entry for EleutherAI's LM Evaluation Harness, an open framework for few-shot evaluation of language models. Gao also co-authored the GPT-NeoX-20B paper, which credits Gao with work on evaluations, the tokenizer, and the section on alignment implications.1234
Work
The PilePublished research
First author of The Pile paper (2020), which describes an 825 GiB English text corpus built from 22 diverse subsets for training large-scale language models.12
LM Evaluation HarnessOpen-source code
First-listed author in the project's citation entry for the evaluation framework.3
GPT-NeoX-20BPublished research
Co-author of the GPT-NeoX-20B paper; its contributions appendix credits Gao with evaluations, tokenizer experiments, and the alignment-implications section.4
Other documented roles
Roles the sources document but that are not confirmed as current.
- Lead author of The Pile (2020), published with an EleutherAI affiliation, EleutherAI2
This profile covers public professional work only, from the sources below. It is not an endorsement or a ranking. To correct or remove it, use the correction form.