YOLOv1 to v8: Unveiling Each Variant–A Comprehensive Review of YOLO
Open accessThe DOI matches its Crossref record; 111 of the 111 references it cites were found in a registry or an index; 13 of the 46 sources its claims cite were checked for support.
Abstract
This paper implements a systematic methodological approach to review the evolution of YOLO variants. Each variant is dissected by examining its internal architectural composition, providing a thorough understanding of its structural components. Subsequently, the review highlights key architectural innovations introduced in each variant, shedding light on the incremental refinements. The review includes benchmarked performance metrics, offering a quantitative measure of each variant’s capabilities. The paper further presents the performance of YOLO variants across a diverse range of domains, manifesting their real-world impact. This structured approach ensures a comprehensive examination of YOLOs journey, methodically communicating its internal advancements and benchmarked performance before delving into domain applications. It is envisioned, the incorporation of concepts such as federated learning can introduce a collaborative training paradigm, where YOLO models benefit from training across multiple edge devices, enhancing privacy, adaptability, and generalisation.
Overview
CoChat Summary
## TL;DR
This paper is a literature review of YOLO from v1 to v8, focusing on architectural evolution, training strategies, benchmark performance, and applications across domains. It also argues that federated learning and edge deployment are promising directions for future YOLO development.
## Key findings
- The review claims to be the first in-depth analysis of mainstream YOLO variants from YOLOv1 to YOLOv8 and says it spans more than 20 domains.
- In the comparative table, reported performance ranges from 63.4% mAP / 45 FPS for YOLOv1 to 69.0% mAP / 52 FPS for YOLOv2, 57.9% mAP / 34 FPS for YOLOv3, 44.3% mAP / 65 FPS for YOLOv4, 50.7% mAP / 200 FPS for YOLOv5, 52.5% mAP / 29 FPS for YOLOv6, 56.8% mAP / 5–160 FPS for YOLOv7, and 53.9% mAP / 280 FPS for YOLOv8.
- YOLOv6 is reported to have eight scaled variants, with the largest variant reaching 57.2% AP on MS COCO test-dev 2017 at about 29 FPS on an NVIDIA Tesla T4.
- Example application results in the survey include 99.7% Accuracy for breast mass detection, 96.42% F1 score for suspicious activity detection, 95.1% MAP for wind turbine blade defect detection, and 92% F1-score for petrochemical pipeline defect detection.
## Methods
- The paper is a systematic review of YOLO variants from YOLOv1 through YOLOv8, organized around architecture, training strategies, performance benchmarks, and real-world applications.
- It draws on benchmark datasets and comparative tables mentioned in the text, especially Pascal VOC and MS COCO; it also discusses dataset diversity and augmentation methods such as Mosaic, MixUp, random affine, and others.
- Main analysis consists of qualitative architectural comparison plus quantitative summary tables of mAP/FPS and application-specific results across domains.
## Limitations
- The paper states that YOLO still faces challenges in occlusions and clutter, scale variations, fine-grained object detection, domain adaptation, biases, and real-time edge deployment.
- The review itself does not report a formal experimental protocol beyond summarizing published results; this is evident from the text because it is a survey and not an original controlled study.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
11 Valid · 20 Not valid · 27 Not checked · 80 Own results and interpretations
The cited sources that could be checked couldn’t confirm or dispute this claim.
In the paper
“Moving away from pre-defined anchor boxes, YOLOv2 extracted anchor boxes using KMEANS [32] clustering, providing enhanced anchor boxes for optimised model training and performance.”
§ YOLOv1 to v8: Unveiling Each Variant -A Comprehensive Review of YOLO > B. YOLOV2
What the sources say
- Could not verifycites [32] matched to “Microsoft COCO: Common Objects in Context” (2014) sourceExperimental — accuracy not yet measured
The Algorithmic Analysis section (pp. 8–9) discusses bounding boxes derived from COCO segmentation masks and evaluates Deformable Parts Models, but does not address YOLOv2’s anchor boxes. Searches for YOLOv2 and k-means anchor-box clustering found no passage in our copy supporting the attributed method. (LLM, conf=0.76)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Could not verify
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“YOLOv3 introduced a foundational generic architecture inspired by the Feature Pyramid Network (FPN) [22], integrating elements such as residual blocks, skip connections, and up-sampling.”
§ YOLOv1 to v8: Unveiling Each Variant -A Comprehensive Review of YOLO > C. YOLOV3
What the sources say
- Partly supportedcites [22] matched to “Feature Pyramid Networks for Object Detection” (2017) sourceExperimental — accuracy not yet measured
Section 3 (p. 3) describes FPN as a general-purpose architecture with a top-down pathway, lateral connections, upsampling, and a ResNet backbone with residual blocks. This supports the identification and description of FPN, but the paper does not establish that YOLOv3 adopted those elements or was inspired by FPN. (LLM, conf=0.86)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
Every source the paper cites for this claim was checked and supports it.
In the paper
“Enthused by the GoogLeNet architecture [28], YOLOv1 deployed a unique approach by substituting GoogLeNet's inception modules with (1 × 1) convolution followed by (3 × 3) convolutional filters.”
§ YOLOv1 to v8: Unveiling Each Variant -A Comprehensive Review of YOLO > III. EVOLUTION OF YOLO ARCHITECTURE: A. YOLOV1
What the sources say
- Supportedcites [28] matched to “Going deeper with convolutions” (2015) source
The citation is attached to “GoogLeNet architecture,” identifying the architecture used as the point of comparison. Section 5 (p. 5) identifies GoogLeNet as an incarnation of the Inception architecture, and Section 5 (p. 6) describes its Inception modules. The citation’s placement does not attribute YOLOv1’s substitution to this source. (LLM, conf=0.90)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“In the realm of autonomous vehicles, YOLO plays a crucial role in object detection for obstacle avoidance and navigation [49].”
§ YOLOv1 to v8: Unveiling Each Variant -A Comprehensive Review of YOLO > VI. REAL-WORLD APPLICATIONS AND IMPACT
What the sources say
- Partly supportedcites [49] matched to “Edge-Computing Video Analytics for Real-Time Traffic Monitoring in a Smart City” (2019) sourceExperimental — accuracy not yet measured
Section 3.2, “Detecting Objects: YOLO V3” (page unavailable), describes YOLO V3 as a real-time object detector used for traffic-flow monitoring. The paper does not address its role in obstacle avoidance or navigation by autonomous vehicles, so that application is not supported by this source. (LLM, conf=0.88)
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Another notable observation showcased in Table 7 is that most variants implemented are v3 onwards.”
§ YOLOv1 to v8: Unveiling Each Variant -A Comprehensive Review of YOLO > Healthcare Imaging and Diagnosis:
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
27 more claims
Create a free account to see all 138 claims and to verify the 9 references still open.
- It may state two separate claims as one. 56
- Probably not a substantial claim. 73
- The quote only partly states this claim. 2
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
“Another limitation is YOLO's computational demands, which hinder its deployment on resource-constrained devices like IoT sensors, mobile platforms, and drones [144].”
From The YOLO Framework: A Comprehensive Review of Evolution, Applications, and Benchmarks in Object Detection (2024) · cites this paper as [144]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ YOLOv1 to v8: Unveiling Each Variant -A Comprehensive Review of YOLO > VIII. CONCLUSION
“| YOLOv6 2021 EfficientRep, CSPDarknet53 | |”
§ YOLOv1 to v8: Unveiling Each Variant -A Comprehensive Review of YOLO > F. YOLOV6
“Meituan Vision AI Department [42] introduced YOLOv6 in September 2022 [43], boasting several innovative features claiming to enhance both efficiency and accuracy.”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ YOLOv1 to v8: Unveiling Each Variant -A Comprehensive Review of YOLO > V. YOLO VERSIONS: A COMPARATIVE ANALYSIS
“YOLOv1: Pioneer in Object Detection (2015) The inaugural version of YOLO, YOLOv1, introduced the groundbreaking concept of real-time object detection using a singlestage architecture with anchor boxes.”
§ YOLOv1 to v8: Unveiling Each Variant -A Comprehensive Review of YOLO > III. EVOLUTION OF YOLO ARCHITECTURE: A. YOLOV1
“Announced in 2016, YOLOv1 marked a profound leap in single-shot object detection.”
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
Very Deep Convolutional Networks for Large-Scale Image Recognition
Simonyan, K., & Zisserman, A. (2014) · arXiv · doi:10.48550/arxiv.1409.1556 · Cited by 114,148
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2017) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2016.2577031 · Cited by 73,472
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Ren, S., He, K., Girshick, R., & Sun, J. (2015) · arXiv · doi:10.48550/arxiv.1506.01497 · Cited by 70,812
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
Going deeper with convolutions
Szegedy, C., Liu, W., Jia, Y., Sermanet, P., Reed, S., Anguelov, D., Erhan, D., Vanhoucke, V., & Rabinovich, A. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298594 · Cited by 48,562
An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., et al. (2020) · arXiv · doi:10.48550/arxiv.2010.11929 · Cited by 47,472
Microsoft COCO: Common Objects in Context
Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., & Zitnick, C. L. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-10602-1_48 · Cited by 45,460
(2016) · Cited by 31,590
Feature Pyramid Networks for Object Detection
Lin, T.-Y., Dollar, P., Girshick, R., He, K., Hariharan, B., & Belongie, S. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.106 · Cited by 30,939
Research works printed but not linked: 1
- Wang, 2022, YOLOv6
Cited by
Items 401–470 of 470
LMM-guided knowledge distillation for power operation object detection in cloud-edge environment
Li, B., Zhang, X., Li, L., Du, J., & Wang, P. (2026) · Journal of Cloud Computing Advances Systems and Applications · doi:10.1186/s13677-026-00872-y
Research on Performance Improvement of Deep Learning Models for Small Target Detection
Qiu, C., Liang, H., & Huang, Q. (2025) · IEEE International Conference on Power, Intelligent Computing and Systems (ICPICS) · doi:10.1109/icpics66386.2025.11347193
A Lightweight Apple Detection Method in Real Orchard Environments Based on Improved YOLOv8s
Zilong, L., & Hongbo, D. (2026) · IEEE Access · doi:10.1109/access.2026.3656532
Zeyi, 张. Z., Zhiyong, 陶. T., Lin, S., & Minglang, 李. L. (2025) · Laser & Optoelectronics Progress · doi:10.3788/lop250894 · Cited by 4
AI-Based Risk Assessment System for Industrial Machine Entrapment Accidents
Jo, Y.-C., Jung, S.-K., Cha, J.-R., et al. (2026) · Journal of the Korean Society of Manufacturing Process Engineers · doi:10.14775/ksmpe.2026.25.8.115
YOLOv8-EVdet: An emergency vehicle detection method for autonomous vehicles
Loukas, P., Bassir, D., & Amanatiadis, A. A. (2026) · International Journal of Advanced Robotic Systems · doi:10.1177/17298806261486243
Rafiq, M. T., Sung, M., Rafiq, G., & Choi, G. S. (2025) · IEEE Access · doi:10.1109/access.2025.3570711 · Cited by 1
Intelligent Fire Detection with UAV-Based Deep Learning for Improved Emergency Response
Vuqar, S., Naile, A., Elmira, T., Bakhriddinova, G., Sadiq, E., Telman, U., & Yaroslava, M. (2026) · Communications in computer and information science · doi:10.1007/978-3-032-39711-9_12
Object Detection Algorithms for Digital Imaging Applications: A Review
Bello, R.-W., Owolawi, P. A., Wyk, E. V., & Tu, C. (2025) · IntechOpen eBooks · doi:10.5772/intechopen.1010205 · Cited by 1
Patil, R. S., & Patil, S. S. (2025) · International Conference for Advancement in Technology (ICONAT) · doi:10.1109/iconat66879.2025.11362689
MarineCrewScenes: A Dataset for Multi-Scale Crew Monitoring in Ship Bridges
Deng, K., & Zhao, X. B. (2025) · International Conference on Computing and Artificial Intelligence (ICCAI) · doi:10.1109/iccai66501.2025.00069
Saleh, H. A., & Isa, N. A. M. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5637958
Alencar, E., & de Freitas, R. (2025) · International Conference on Soft Computing & Machine Intelligence (ISCMI) · doi:10.1109/iscmi67495.2025.11358466
EMSH-DETR: an efficient multi-scale and hybrid DETR for floating garbage detection
Liu, C., Li, J., Ke, Z., Yang, X., Hu, C., & Zou, T. (2025) · Measurement Science and Technology · doi:10.1088/1361-6501/ae2b2c · Cited by 1
Cheranchery, M. F., Ponnan, B., Shihabudheen, K. V., Navandar, Y. V., Faouzi, N. E., & Furno, A. (2026) · Innovative Infrastructure Solutions · doi:10.1007/s41062-026-02967-6
Detection Of Pufferfish Using Computer Vision And Deep Learning Methods
Yüksel, H. U., & Tonguç, G. (2025) · El-Cezeri Fen ve Mühendislik Dergisi · doi:10.31202/ecjse.1628790 · Cited by 1
A Unified Pipeline for Robust Shuttlecock Tracking and Court-Space Visualization
Ngo, K. A., Nguyen, A. Q., Huynh, T. K., et al. (2026) · International Conference on Intelligent Information Technology (ICIIT) · doi:10.1145/3805862.3805945
Lin, H., & Lin, C.-L. (2026) · Expert Systems with Applications · doi:10.1016/j.eswa.2026.134369
Evaluating Real-Time Object Detection Models for Autonomous Vehicle Vision
Bahri, S., & Mouftah, H. T. (2025) · IEEE World AI IoT Congress (AIIoT) · doi:10.1109/aiiot65859.2025.11105268
Cabral, P. S., da Silva, J. C., Diniz, J. B., & Gomes, D. L. (2025) · doi:10.5753/sbcas_estendido.2025.7108
Application and Optimization of YOLOv11 for Traffic Object Detection
Dai, Z., Lu, Y., & Wu, R. (2025) · IEEE Smart World Congress (SWC) · doi:10.1109/swc65939.2025.00191
CatCare – Intelligent IoP System for Recognition, Monitoring and Surveillance of Cats in Colonies
Fernandim, L., Gonçalves, C., Caetano, F., et al. (2026) · Lecture notes in networks and systems · doi:10.1007/978-3-032-10929-3_40
Real Time Detection Network for Multi-Scale Targets in Remote Sensing Images
Qu, J., Yang, R., & Lin, Y. (2025) · International Conference on Machine Learning, Big Data and Business Intelligence (MLBDBI) · doi:10.1109/mlbdbi67855.2025.11331317
Edge-Aware Multi-Sensor YOLOV12: Lightweight Real-Time Object Detection for Smart Transportation
Ma, L., & Ji, L. (2026) · Lecture notes in electrical engineering · doi:10.1007/978-981-95-8232-7_23
Breast Cancer Classification: A Semantic Comparison of YOLOv8 and YOLOv11
Singh, P., & Kumar, S. (2025) · Lecture notes in networks and systems · doi:10.1007/978-981-96-7134-2_26
Liu, G., Huang, Y., Wu, H., & Zhang, H. (2026) · Land · doi:10.3390/land15071119 · Cited by 1
Zhang, H., Chen, Z., Huang, Y., et al. (2026) · Nondestructive Testing And Evaluation · doi:10.1080/10589759.2026.2737805
Lightweight UAV-Based Road Disease Detection Algorithm
Jinbo, G., Wang, S., Fenghua, X., et al. (2026) · Lecture notes in electrical engineering · doi:10.1007/978-981-95-1802-9_14
Data Augmentation for YOLO Based Road Object Detection in Autonomous Vehicles
Omodaratan, B., Obafemi, O. O., & Adedeji, P. A. (2025) · IEEE Connecting the Unconnected Europe, Middle East, and Africa Regional Summit (CTUS-EMEA) · doi:10.1109/ctus-emea68989.2025.11592192
YOLOv5-TS: Channel-Spatial Attention Guided Bidirectional Feature Fusion for Traffic Sign Detection
Zhang, Y., Zhang, F., Gao, Y., et al. (2025) · Communications in computer and information science · doi:10.1007/978-3-031-99879-9_6
LSO-YOLO: a lightweight real-time object detection network for UAV small-object detection
Dai, P., Chen, L., Lian, Y., & Xie, L. (2026) · Journal of Real-Time Image Processing · doi:10.1007/s11554-026-01977-y
Performance Analysis of Faster R-CNN and YOLOv8 Model for Mango Fruits Detection
Abdullah, M. H. L., & Eldeib, M. A. (2026) · Journal of Informatics and Web Engineering · doi:10.33093/jiwe.2026.5.2.17
Reyes, H. E. Z., Barreto, H. P., & López-Armas, G. D. C. (2026) · Mathematics · doi:10.3390/math14122080 · Cited by 1
Evaluating Real-Time Object Detection Models for Autonomous Vehicular Vision applications
Bahri, S., & Mouftah, H. T. (2025) · International Wireless Communications and Mobile Computing Conference (IWCMC) · doi:10.1109/iwcmc65282.2025.11059704
Rustam, S. A., Khan, S. N., & Khan, S. (2026) · International Journal of Innovations in Science and Technology · doi:10.33411/ijist/1807
A spiking neural network for energy-efficient power transmission line inspection
Rong, H., Bi, T., Zhang, G., & Neri, F. (2026) · Computer-Aided Civil and Infrastructure Engineering · doi:10.1016/j.cacaie.2026.100042 · Cited by 1
Xin, L., & Li, Y. (2025) · Chinese Control Conference (CCC) · doi:10.23919/ccc64809.2025.11178470
Revathi, P., Baluprithviraj, K. N., R, D. V., Hemadharshini, S. V., & Kavin Kumar, P. (2024) · International Conference on Sustainable Communication Networks and Application (ICSCNA) · doi:10.1109/icscna63714.2024.10864334
Mustapha, H., Wang, Z., Mehedi, I. M., Wu, Z., & Zhang, S. (2026) · International Conference on Computer Vision and Data Mining (ICCVDM) · doi:10.1109/iccvdm71812.2026.11709366
Aeri, M., RAYAT, S. S., & Thapa, S. (2026) · Research Square · doi:10.21203/rs.3.rs-8912354/v1
A real- time spatial occupancy based framework for predicting green signal time
Nigam, N., Gupta, S. K., Singh, D. P., Choudhary, J., Haripriya, R., & Priya, A. (2026) · Discover Computing · doi:10.1007/s10791-026-10506-1 · Cited by 1
Ground‐based robotic remote sensing for standardized biodiversity monitoring in coastal habitats
Lorenzo, G. D., Bagella, S., Rasino, M. D. V., Carranza, M. L., Garabini, M., & Angelini, F. (2026) · Remote Sensing in Ecology and Conservation · doi:10.1002/rse2.70074
Susanto, A., Sumpeno, S., & Purnomo, M. H. (2026) · International Electronics Symposium (IES) · doi:10.1109/ies71033.2026.11702628
Onboard Vehicle Detection for Autonomous UAV Reconnaissance under Degraded Visual Conditions
Pham, T. D., & Тарлаковский, Д. В. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-34384-0_28
A precise deep learning approach for timber traceability along the forest value chain
Sandvik, Y. J., Hoseini, M., & Fischer, C. (2025) · International Journal of Forest Engineering · doi:10.1080/14942119.2025.2572938
Metadata
- Authors
- Hussain, M.
- Affiliations
- University of Huddersfield (GB)
- Corresponding author
- Hussain, Muhammad
- ORCID
- Hussain, Muhammad
- Venue
- IEEE Access
- Published
- 2024
- Volume
- 12
- Pages
- 42816-42833
- ISSN
- 2169-3536 (ISSN-L)
- OpenAlex
- W4392939809
- Semantic Scholar
- 6ccd0eb94f37c00ae49de81b2752dbb9bd6b0c3b
- Cited by
- 492
- References
- 105
- Citations by year
- 2026: 168 · 2025: 253 · 2024: 71
- FWCI
- 68.53
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- IEEE Access (publishedVersion, cc-by-nc-nd)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Huddersfield Research Portal (University of Huddersfield) (submittedVersion, cc-by-nc-nd)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Life Sciences › Agricultural and Biological Sciences › Plant Science › Smart Agriculture and AI
- Keywords
- YOLO, object detection, YOLOv8, benchmark performance, federated learning, edge devices, privacy-preserving training
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.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.