Home·Curated Catalog·NLP / Text
💬 Curated Catalog · NLP / Text

TriviaQA — Large-Scale Reading Comprehension

Trivia questions paired with Wikipedia + web evidence — long-form reading comprehension at scale.

LQS 78 · gold ✓ Commercial OK 650K question-answer-evidence triples 2.4 GB JSON · JSONL Released 2017
Browse commercial NLP / Text → Visit original source ↗
Source: nlp.cs.washington.edu · maintained by University of Washington (AI2)
650K
question-answer-evidence triples
2.4 GB
Size on disk
78
LQS · gold
2017
First released

About this dataset

TriviaQA is a large-scale reading comprehension dataset containing 650K question-answer-evidence triples. Questions are authored by trivia enthusiasts (higher syntactic complexity than crowd-sourced QA), and each is paired with both Wikipedia articles and web search results as evidence documents. Distinct from SQuAD-style benchmarks because answers can require multi-sentence or cross-document reasoning.

Maintainer
University of Washington (AI2)
License
Apache 2.0
Formats
JSON · JSONL
Paper
Read on arxiv.org →

LabelSets Quality Score

LQS is our 7-dimension quality score, computed from the dataset's published statistics. See methodology →

78
out of 100
gold tier

Solid dataset with some trade-offs

Composite score computed from the 7 dimensions below: completeness, uniqueness, validation health, size adequacy, format compliance, label density, and class balance.

Completeness 92
No public completeness metric; using prior for 'research_release' datasets.
Uniqueness 68
Minimal deduplication disclosed.
Validation 68
Crowdsourced labels without disclosed QC protocol.
Size adequacy 92
650,000 items — exceeds 100,000 adequacy target for NLP / Text.
Format compliance 95
Industry-standard format — drop-in compatible with mainstream tooling.
Label density 52
Average 1.0 labels per item (sparse).
Class balance 75
Moderate class skew — realistic production distribution.

What it's used for

Common tasks and benchmarks where TriviaQA — Large-Scale Reading Comprehension is the default or competitive choice.

Sample statistics

What's actually in the dataset — from the maintainer's published stats.

650K QA pairs, 95K unique questions, ~6 evidence documents per question on average.

License

TriviaQA — Large-Scale Reading Comprehension is distributed under Apache 2.0. This is a third-party public dataset; LabelSets indexes and scores it but does not host or redistribute the data. Always verify current license terms with the maintainer before commercial use.

Need commercial-licensed NLP / Text data?

LabelSets sellers offer paid nlp / text datasets with what public datasets often can't give you:

Browse paid NLP / Text → Sell your dataset

Similar public datasets

Other entries in the NLP / Text catalog.

SQuAD 2.0 — Stanford Question Answering Dataset
150K Q&A pairs · CC BY-SA 4.0
The Pile
300M documents · MIT (mixed sub-licenses)
C4 — Colossal Clean Crawled Corpus
365M documents · ODC-BY
Wikipedia (English Dump)
6.8M articles · CC BY-SA 4.0

Frequently Asked Questions

TriviaQA — Large-Scale Reading Comprehension is distributed under Apache 2.0, which generally permits commercial use. Always verify the current license terms with the maintainer (University of Washington (AI2)) before using in a commercial product.
TriviaQA — Large-Scale Reading Comprehension contains 650,000 question-answer-evidence triples. 650K QA pairs, 95K unique questions, ~6 evidence documents per question on average.
TriviaQA — Large-Scale Reading Comprehension is maintained by University of Washington (AI2) and is available at https://nlp.cs.washington.edu/triviaqa/. LabelSets indexes and scores this dataset for discoverability but does not redistribute it.
LQS is a 7-dimension quality score (completeness, uniqueness, validation, size adequacy, format compliance, label density, class balance) computed from the dataset's published statistics. Composite scores map to tiers: platinum (≥90), gold (≥75), silver (≥60), bronze (<60). Read the full methodology.