NLP Evaluation in trouble: On the Need to Measure LLM Data Contamination for each Benchmark

Oscar Sainz, Jon Campos, Iker García-Ferrero, Julen Etxaniz, Oier Lopez de Lacalle, Eneko Agirre


Abstract
In this position paper we argue that the classical evaluation on Natural Language Processing (NLP) tasks using annotated benchmarks is in trouble. The worst kind of data contamination happens when a Large Language Model (LLM) is trained on the test split of a benchmark, and then evaluated in the same benchmark. The extent of the problem is unknown, as it is not straightforward to measure. Contamination causes an overestimation of the performance of a contaminated model in a target benchmark and associated task with respect to their non-contaminated counterparts. The consequences can be very harmful, with wrong scientific conclusions being published while other correct ones are discarded. This position paper defines different levels of data contamination and argues for a community effort, including the development of automatic and semi-automatic measures to detect when data from a benchmark was exposed to a model, and suggestions for flagging papers with conclusions that are compromised by data contamination.
Anthology ID:
2023.findings-emnlp.722
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2023
Month:
December
Year:
2023
Address:
Singapore
Editors:
Houda Bouamor, Juan Pino, Kalika Bali
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
10776–10787
Language:
URL:
https://aclanthology.org/2023.findings-emnlp.722
DOI:
10.18653/v1/2023.findings-emnlp.722
Bibkey:
Cite (ACL):
Oscar Sainz, Jon Campos, Iker García-Ferrero, Julen Etxaniz, Oier Lopez de Lacalle, and Eneko Agirre. 2023. NLP Evaluation in trouble: On the Need to Measure LLM Data Contamination for each Benchmark. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 10776–10787, Singapore. Association for Computational Linguistics.
Cite (Informal):
NLP Evaluation in trouble: On the Need to Measure LLM Data Contamination for each Benchmark (Sainz et al., Findings 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.findings-emnlp.722.pdf