The DOI matches its Crossref record; 27 of the 27 references it cites were found in a registry or an index.
SQuAD: 100,000+ Questions for Machine Comprehension of Text
Open accessAbstract
We present the Stanford Question Answering Dataset (SQuAD), a new reading comprehension dataset consisting of 100,000+ questions posed by crowdworkers on a set of Wikipedia articles, where the answer to each question is a segment of text from the corresponding reading passage. We analyze the dataset to understand the types of reasoning required to answer the questions, leaning heavily on dependency and constituency trees. We build a strong logistic regression model, which achieves an F1 score of 51.0%, a significant improvement over a simple baseline (20%). However, human performance (86.8%) is much higher, indicating that the dataset presents a good challenge problem for future research. The dataset is freely available at https://stanford-qa.com
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Both metrics assume short, span-extractable answers, an assumption fundamentally violated by Think-class models that produce reasoning traces often running to several hundred tokens.”
From When Context Overwhelms: A Study of Long-Context vs. Retrieval-Based QA Under Noise (2026) · cites this paper as [15]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“We use SQuAD 1.1 which excludes the unanswerable questions added in v.2.0 that are deliberately constructed to induce mistakes so they do not in practice cause confabulations to occur.”
From Detecting hallucinations in large language models using semantic entropy (2024) · cites this paper as [33]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The top-3 softmax-weighted configuration trades 1 . 6 F1 for multi-chunk fusion”
From Structured Memory for Edge Language Models: Persistent Context and Corpus Retrieval via O(1) SSM State Injection (2026) · cites this paper as [33]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The empirically optimal k is therefore corpus-dependent: k =1 for single-paragraph evidence (SQuAD), k =2 -3 for multi-hop (HotpotQA).”
From Structured Memory for Edge Language Models: Persistent Context and Corpus Retrieval via O(1) SSM State Injection (2026) · cites this paper as [33]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Because we use the released SQuAD reference answers verbatim and do not introduce model-assisted question generation, no additional human verification step is required.”
From Structured Memory for Edge Language Models: Persistent Context and Corpus Retrieval via O(1) SSM State Injection (2026) · cites this paper as [33]
Experimental — accuracy not yet measured
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“We report token-level Exact Match (EM) and F1 against the gold answer set, computed via the official SQuAD v1.1 evaluation script. 1 The script applies the standard SQuAD normalizationlowercasing, stripping of punctuation, removal of articles ( a , an , the ), and whitespace tokenization-to both predictions and references before scoring”
From Structured Memory for Edge Language Models: Persistent Context and Corpus Retrieval via O(1) SSM State Injection (2026) · cites this paper as [33]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“We use SQuAD (Rajpurkar et al., 2016) and HotpotQA (Yang et al., 2018) as the base datasets, which are two widely used datasets in the field of QA and QG.”
From QGEval: Benchmarking Multi-dimensional Evaluation for Question Generation (2024) · cites this paper as Rajpurkar et al., 2016
The cited source says what the claim attributes to it.
In the citing paper
“Our results show that, following reinforcement learning (RL) training, finetuning on self-generated synthetic data improves question-answering performance on the no-passage-in-context variant of SQuAD [13] from 33.5% to 47.0%.”
From Self-Adapting Language Models (2025) · cites this paper as [13]
The cited source says what the claim attributes to it.
In the citing paper
“Standard NLP models are often evaluated on large datasets of coherent text-based questions and answers [48, 49] or paired written dialogue [68].”
From A Mixed-Methods Approach to Understanding User Trust after Voice Assistant Failures (2023) · cites this paper as [49]
The cited source says what the claim attributes to it.
In the citing paper
“We evaluate on a randomly sampled 1 , 000 -question subset of the SQuAD v1.1 dev split using the official evaluation script”
From Structured Memory for Edge Language Models: Persistent Context and Corpus Retrieval via O(1) SSM State Injection (2026) · cites this paper as [33]
The cited source says what the claim attributes to it.
In the citing paper
“For tractability under the 19 tok/s edge-throughput inference configuration, we sample a fixed random subset of 1 , 000 questions from the SQuAD v1.1 dev split using a fixed random seed.”
From Structured Memory for Edge Language Models: Persistent Context and Corpus Retrieval via O(1) SSM State Injection (2026) · cites this paper as [33]
The cited source says what the claim attributes to it.
In the citing paper
“Traditional question-answering (QA) benchmarks are typically designed to provide fixed answers to questions, as in datasets like SQuAD (Rajpurkar et al., 2016) and Natural Questions (Kwiatkowski et al., 2019).”
From DebateQA: Evaluating Question Answering on Debatable Knowledge (2026) · cites this paper as Rajpurkar et al., 2016
The check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“We found that existing NLP datasets do not cover the wide breadth of possible conversational failure cases due to their emphasis on correct data for training.”
From A Mixed-Methods Approach to Understanding User Trust after Voice Assistant Failures (2023) · cites this paper as [49]
Experimental — accuracy not yet measuredThe check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“PRECOG top-1 matches in-context RAG to within 0 . 2 F1 and 0 . 2 EM on this subset”
From Structured Memory for Edge Language Models: Persistent Context and Corpus Retrieval via O(1) SSM State Injection (2026) · cites this paper as [33]
Experimental — accuracy not yet measuredThe check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“The total chunk count is 2 , 067 , producing a state corpus of approximately 0 . 40 GB; the corresponding sentence-encoder key index occupies under 1 MB.”
From Structured Memory for Edge Language Models: Persistent Context and Corpus Retrieval via O(1) SSM State Injection (2026) · cites this paper as [33]
Experimental — accuracy not yet measuredIn the citing paper
“The type of questions to be generated range from the typical short answer span questions to multiple spans for question generation from the text [42, 47, 106, 107, 112].”
From Automatic question generation: a review of methodologies, datasets, evaluation metrics, and applications (2023) · cites this paper as [42]
In the citing paper
“Most datasets that are currently available for training question generation systems are crowd-sourced [42, 47, 84, 92, 114, 115, 119, 120, 122, 123], which largely impacts the quality of generated questions.”
From Automatic question generation: a review of methodologies, datasets, evaluation metrics, and applications (2023) · cites this paper as [42]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
ImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333
Building a Large Annotated Corpus of English: The Penn Treebank
Marcus, M. (1993) · doi:10.21236/ada273556 · Cited by 7,405
Teaching Machines to Read and ComprehendPreprint
Hermann, K. M., Kočiský, T., Grefenstette, E., Espeholt, L., Kay, W., Suleyman, M., & Blunsom, P. (2015) · arXiv (Cornell University) · doi:10.48550/arxiv.1506.03340 · Cited by 2,855
Building Watson: An Overview of the DeepQA Project
Ferrucci, D. A., Brown, E. W., Chu‐Carroll, J., Fan, J., Gondek, D. C., Kalyanpur, A. A., Lally, A., Murdock, J. W., Nyberg, E., Prager, J. M., et al. (2010) · AI Magazine · doi:10.1609/aimag.v31i3.2303 · Cited by 1,523
WikiQA: A Challenge Dataset for Open-Domain Question Answering
Yang, Y., Yih, W.-T., & Meek, C. (2015) · Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/d15-1237 · Cited by 829
Learning surface text patterns for a Question Answering system
Ravichandran, D., & Hovy, E. H. (2001) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.3115/1073083.1073092 · Cited by 815
MCTest: A Challenge Dataset for the Open-Domain Machine Comprehension of Text
Richardson, M., Burges, C., & Renshaw, E. (2013) · Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/d13-1020 · Cited by 778
Building a question answering test collection
Voorhees, E., & Tice, D. M. (2000) · Annual International ACM SIGIR Conference on Research and Development in Information Retrieval · doi:10.1145/345508.345577 · Cited by 670
Towards AI-Complete Question Answering: A Set of Prerequisite Toy Tasks
Weston, J., Bordes, A., Chopra, S., Rush, A. M., van Merriënboer, B., Joulin, A., & Mikolov, T. (2015) · arXiv · doi:10.48550/arxiv.1502.05698 · Cited by 670
A Thorough Examination of the CNN/Daily Mail Reading Comprehension Task
Chen, D., Bolton, J., & Manning, C. D. (2016) · Proceedings of the conference - Association for Computational Linguistics. Meeting · doi:10.18653/v1/p16-1223 · Cited by 494
Cited by
Sorted by citations
Items 1,101–1,109 of 1,109
Gopal, A. M., Pirbadian, A., Carlson, K. D., Lewis, M. A., & Tapson, J. (2026) · arXiv
“1 development set [33].”
ATLAS: Learning to Optimally Memorize the Context at Test Time
Behrouz, A., Li, Z., Kacham, P., Daliri, M., Deng, Y., Zhong, P., Razaviyayn, M., & Mirrokni, V. (2025) · arXiv · doi:10.48550/arxiv.2505.23735
When Context Overwhelms: A Study of Long-Context vs. Retrieval-Based QA Under Noise
Akbar, Q. U. N., & Akbar, Q. M. (2026) · doi:10.33774/coe-2026-tvxc6
Signal and Noise: A Framework for Reducing Uncertainty in Language Model Evaluation
Heineman, D., Hofmann, V., Magnusson, I., Gu, Y., Smith, N. A., Hajishirzi, H., Lo, K., & Dodge, J. (2025) · arXiv · doi:10.48550/arxiv.2508.13144
YESciEval: Robust LLM-as-a-Judge for Scientific Question Answering
D'Souza, J., Giglou, H. B., & Münch, Q. (2025) · arXiv · doi:10.48550/arxiv.2505.14279
Chen, Y., Zhang, C., Luo, D., D'Haro, L. F., Tan, R. T., & Li, H. (2024) · arXiv · doi:10.48550/arxiv.2405.14646
A Dataset of Information-Seeking Questions and Answers Anchored in Research Papers
Dasigi, P., Lo, K., Beltagy, I., Cohan, A., Smith, N. A., & Gardner, M. (2021) · arXiv · doi:10.48550/arxiv.2105.03011
Mem-α: Learning Memory Construction via Reinforcement Learning
Wang, Y., Takanobu, R., Liang, Z., Mao, Y., Hu, Y., McAuley, J., & Wu, X. (2025) · arXiv · doi:10.48550/arxiv.2509.25911
In ChatGPT We Trust? Measuring and Characterizing the Reliability of ChatGPT
Shen, X., Chen, Z., Backes, M., & Zhang, Y. (2023) · arXiv · doi:10.48550/arxiv.2304.08979 · Cited by 76
“We employ 10 widely used benchmark QA datasets in our study, including BoolQ [24], OpenbookQA (OQA) [51], RACE [46], ARC [25], CommonsenseQA (CQA) [67], SQuAD1 [60], SQuAD2 [59], NarrativeQA (NQA) [42], ELI5 [28], and TruthfulQA (TQA) [50].”
Influential
Metadata
- Authors
- Rajpurkar, P., Zhang, J., Lopyrev, K., & Liang, P.
- Affiliations
- Stanford University (US)
- Venue
- Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing
- Published
- 2016
- Pages
- 2383-2392
- OpenAlex
- W2963748441
- Cited by
- 6,208
- References
- 27
- Citations by year
- 2026: 335 · 2025: 492 · 2024: 506 · 2023: 1093 · 2022: 879 · 2021: 1127 · 2020: 858 · 2019: 592 · 2018: 254 · 2017: 62 · 2016: 3
- FWCI
- 163.81
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Conference on Empirical Methods in Natural Language Processing (EMNLP) (publishedVersion, cc-by)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Physical Sciences › Computer Science › Information Systems › Information Retrieval and Search Behavior
- Keywords
- question answering, reading comprehension, SQuAD, Wikipedia, dependency trees, constituency trees, logistic regression, human performance comparison, natural language understanding
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
