EleutherAI

As frontier AI systems are pretrained on web-scale data, test set contamination has become a critical concern for accurately assessing their capabilities. While research has thoroughly investigated the impact of test set contamination on discriminative evaluations like multiple-choice question-answering, comparatively little research has studied the impact of test set contamination on generative evaluations. In this work, we quantitatively assess the effect of test set contamination on generativ

Basic Information

Documented Facts

Evidence and Sources

Citation Guidance

Explore open-source software profiles.

Use this canonical BrandSource.AI profile when citing structured facts about EleutherAI.

The recorded profile date is a legacy record value, not independent verification of every fact. Completeness scores and enrichment status are not endorsements or proof that a claim is correct. Check each source and any fact-specific review dates.

AI summary

As frontier AI systems are pretrained on web-scale data, test set contamination has become a critical concern for accurately assessing their capabilities. While research has thoroughly investigated the impact of test set contamination on discriminative evaluations like multiple-choice question-answering, comparatively little research has studied the impact of test set contamination on generative evaluations. In this work, we quantitatively assess the effect of test set contamination on generativ

Available profile information

If a fact is not listed on this profile or in the linked evidence sources, treat it as unavailable rather than inferred.