Depending on the amount of data to process, file generation may take longer.

If it takes too long to generate, you can limit the data by, for example, reducing the range of years.

Chapter

Download BibTeX

Title

Leak, Cheat, Repeat: Data Contamination and Evaluation Malpractices in Closed-Source LLMs

Authors

[ 1 ] Instytut Informatyki, Wydział Informatyki i Telekomunikacji, Politechnika Poznańska | [ P ] employee

Scientific discipline (Law 2.0)

[2.3] Information and communication technology

Year of publication

2024

Chapter type

chapter in monograph / paper

Publication language

english

Abstract

EN Natural Language Processing (NLP) research is increasingly focusing on the use of Large Language Models (LLMs), with some of the most popular ones being either fully or partially closed-source. The lack of access to model details, especially regarding training data, has repeatedly raised concerns about data contamination among researchers. Several attempts have been made to address this issue, but they are limited to anecdotal evidence and trial and error. Additionally, they overlook the problem of indirect data leaking, where modelsare iteratively improved by using data coming from users. In this work, we conduct the first systematic analysis of work using OpenAI’s GPT-3.5 and GPT-4, the most prominently used LLMs today, in the context of data contamination. By analysing 255 papers and considering OpenAI’s data usage policy, we extensively document the amount of data leaked to these models during the first year after the model’s release. We report that these models have been globally exposed to ∼4.7M samples from 263 benchmarks. At the same time, we document a number of evaluation malpractices emerging in the reviewed papers, such as unfair or missing baseline comparisons and reproducibility issues. We release our results as a collaborative project on https://leak-llm.github.io/, where other researchers can contribute to our efforts.

Pages (from - to)

67 - 93

URL

https://aclanthology.org/2024.eacl-long.5/

Book

The 18th Conference of the European Chapter of the Association for Computational Linguistics EACL 2024 : Proceedings of the Conference, Vol. 1 (Long Papers)

Presented on

18th Conference of the European Chapter of the Association for Computational Linguistics EACL 2024, 17-22.03.2024, St. Julians, Malta

License type

CC BY (attribution alone)

Open Access Mode

publisher's website

Open Access Text Version

final published version

Date of Open Access to the publication

at the time of publication

Ministry points / chapter

20

Ministry points / conference (CORE)

140

This website uses cookies to remember the authenticated session of the user. For more information, read about Cookies and Privacy Policy.