WikiText

WikiText scores raw language-modelling quality by perplexity on curated Wikipedia articles; it measures how well a model predicts text, not whether it answers a task correctly.

Also known as: WikiText-2, WikiText-103

unassessed

This page is a discovery lead. Nobody has yet assessed it against the catalogue contract, so it carries no disposition. Absence of evidence here is not evidence of staleness.
Categorygeneration
Subcategorylanguage-modelling perplexity over curated Wikipedia text
Page statusactive
Metricperplexity (word-level and/or byte-level; bits-per-byte is also reported by some harnesses)
Directionlower_is_better
Unitperplexity
Dataset licenceCC BY-SA 3.0 and GFDL (dual-licensed, per the Hugging Face dataset card; inherited from Wikipedia's own article licensing)
PublisherMetaMind, a Salesforce company (now Salesforce Research)

What it measures

WikiText is not a task benchmark: it gives a model a stream of curated Wikipedia article text and measures how well the model's predicted next-token probabilities match what actually comes next, scored as perplexity. There are no questions, answers or instructions to follow, so a WikiText score says nothing directly about a model's ability to solve a problem, follow an instruction or reason -- it measures modelling quality, the calibration and fluency of the model's underlying probability distribution over English text, which is a different and narrower thing than the accuracy percentage reported by a question-answering or reasoning benchmark.

Task format

A model reads long-form Wikipedia article text (WikiText-2 or the ~110x larger WikiText-103) and, for each token, produces a probability distribution over the vocabulary; no prompt, instruction or answer is generated. Evaluation is typically "rolling" loglikelihood scoring across each document, letting a model use as much preceding context as its window allows rather than scoring fixed, isolated chunks.

Models reporting this benchmark

No model card in ModelSpec reports this benchmark yet.

Data

This page as JSON · Edit on GitHub