Eyeriss: a spatial architecture for energy-efficient dataflow for convolutional neural networks
The DOI matches its Crossref record; 45 of the 48 references it cites were found in a registry or an index.
Abstract
Deep convolutional neural networks (CNNs) are widely used in modern AI systems for their superior accuracy but at the cost of high computational complexity. The complexity comes from the need to simultaneously process hundreds of filters and channels in the high-dimensional convolutions, which involve a significant amount of data movement. Although highly-parallel compute paradigms, such as SIMD/SIMT, effectively address the computation requirement to achieve high throughput, energy consumption still remains high as data movement can be more expensive than computation. Accordingly, finding a dataflow that supports parallel processing with minimal data movement cost is crucial to achieving energy-efficient CNN processing without compromising accuracy. In this paper, we present a novel dataflow, called row-stationary (RS), that minimizes data movement energy consumption on a spatial architecture. This is realized by exploiting local data reuse of filter weights and feature map pixels, i.e., activations, in the high-dimensional convolutions, and minimizing data movement of partial sum accumulations. Unlike dataflows used in existing designs, which only reduce certain types of data movement, the proposed RS dataflow can adapt to different CNN shape configurations and reduces all types of data movement through maximally utilizing the processing engine (PE) local storage, direct inter-PE communication and spatial parallelism. To evaluate the energy efficiency of the different dataflows, we propose an analysis framework that compares energy cost under the same hardware area and processing parallelism constraints. Experiments using the CNN configurations of AlexNet show that the proposed RS dataflow is more energy efficient than existing dataflows in both convolutional (1.4× to 2.5×) and fully-connected layers (at least 1.3× for batch size larger than 16). The RS dataflow has also been demonstrated on a fabricated chip, which verifies our energy analysis.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
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 · cited 1×
“The recent popularity of deep learning [1], specifically deep convolutional neural networks (CNNs), can be attributed to its ability to achieve unprecedented accuracy for tasks ranging from object recognition [2] [3] [4] [5] and detection [6] [7] to scene understanding [8].”
BackgroundImageNet classification with deep convolutional neural networks
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2017) · Communications of the ACM · doi:10.1145/3065386 · Cited by 112,155 · cited 1×
“The recent popularity of deep learning [1], specifically deep convolutional neural networks (CNNs), can be attributed to its ability to achieve unprecedented accuracy for tasks ranging from object recognition [2] [3] [4] [5] and detection [6] [7] to scene understanding [8].”
BackgroundLeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
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
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Girshick, R., Donahue, J., Darrell, T., & Malik, J. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2014.81 · Cited by 32,684
(2010) · International Conference on Machine Learning · Cited by 13,995
Caffe: Convolutional Architecture for Fast Feature Embedding
Jia, Y., Shelhamer, E., Donahue, J., Karayev, S., Long, J., Girshick, R., Guadarrama, S., & Darrell, T. (2014) · Proceedings of the 22nd ACM international conference on Multimedia · doi:10.1145/2647868.2654889 · Cited by 11,077 · cited 2×
“Table II CONV/FC LAYER SHAPE CONFIGURATIONS IN ALEXNET [39].”
BackgroundEyeriss: An Energy-Efficient Reconfigurable Accelerator for Deep Convolutional Neural Networks
Chen, Y.-H., Krishna, T., Emer, J. S., & Sze, V. (2017) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2016.2616357 · Cited by 3,247 · cited 3×
“The RS dataflow is a key feature of the Eyeriss architecture, which has been implemented in a fabricated chip [41] (Fig.”
BackgroundMethodInfluentialSelf-citationOverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks
Sermanet, P., Eigen, D., Zhang, X., Mathieu, M., Fergus, R., & LeCun, Y. (2013) · arXiv · doi:10.48550/arxiv.1312.6229 · Cited by 2,680 · cited 1×
“The recent popularity of deep learning [1], specifically deep convolutional neural networks (CNNs), can be attributed to its ability to achieve unprecedented accuracy for tasks ranging from object recognition [2] [3] [4] [5] and detection [6] [7] to scene understanding [8].”
Background(2014) · DSpace@MIT (Massachusetts Institute of Technology) · Cited by 2,534
Convolutional networks and applications in vision
LeCun, Y., Kavukcuoglu, K., & Farabet, C. (2010) · Proceedings of 2010 IEEE International Symposium on Circuits and Systems · doi:10.1109/iscas.2010.5537907 · Cited by 2,213 · cited 1×
“A convolutional neural network (CNN) is constructed by stacking multiple computation layers as a directed acyclic graph [36].”
MethodOptimizing FPGA-based Accelerator Design for Deep Convolutional Neural Networks
Zhang, C., Li, P., Sun, G., Guan, Y., Xiao, B., & Cong, J. (2015) · Proceedings of the 2015 ACM/SIGDA International Symposium on Field-Programmable Gate Arrays · doi:10.1145/2684746.2689060 · Cited by 2,158 · cited 1×
“Many previous papers have proposed specialized CNN dataflows on various platforms, including GPU [14], FPGA [15] [16] [17] [18] [19] [20] [21], and ASIC [22] [23] [24] [25] [26].”
Background1.1 Computing's energy problem (and what we can do about it)
Horowitz, M. (2014) · 2014 IEEE International Solid-State Circuits Conference Digest of Technical Papers (ISSCC) · doi:10.1109/isscc.2014.6757323 · Cited by 2,069
Deep Learning with Limited Numerical Precision
Gupta, S., Agrawal, A., Gopalakrishnan, K., & Narayanan, P. (2015) · International Conference on Machine Learning · Cited by 1,590
DaDianNao: A Machine-Learning Supercomputer
Chen, Y., Luo, T., Liu, S., Zhang, S., He, L., Wang, J., Li, L., Chen, T., Xu, Z., Sun, N., et al. (2014) · 2014 47th Annual IEEE/ACM International Symposium on Microarchitecture · doi:10.1109/micro.2014.58 · Cited by 1,492 · cited 1×
“Many previous papers have proposed specialized CNN dataflows on various platforms, including GPU [14], FPGA [15] [16] [17] [18] [19] [20] [21], and ASIC [22] [23] [24] [25] [26].”
BackgroundSelf-citationDeep Learning with Limited Numerical Precision
Gupta, S., Agrawal, A., Gopalakrishnan, K., & Narayanan, P. (2015) · arXiv · doi:10.48550/arxiv.1502.02551 · Cited by 1,364 · cited 1×
“Many previous papers have proposed specialized CNN dataflows on various platforms, including GPU [14], FPGA [15] [16] [17] [18] [19] [20] [21], and ASIC [22] [23] [24] [25] [26].”
BackgroundDianNao: a small-footprint high-throughput accelerator for ubiquitous machine-learning
Chen, T., Du, Z., Sun, N., Wang, J., Wu, C., Chen, Y., & Temam, O. (2014) · Proceedings of the 19th international conference on Architectural support for programming languages and operating systems · doi:10.1145/2541940.2541967 · Cited by 1,297 · cited 2×
“Many previous papers have proposed specialized CNN dataflows on various platforms, including GPU [14], FPGA [15] [16] [17] [18] [19] [20] [21], and ASIC [22] [23] [24] [25] [26].”
BackgroundSelf-citationcuDNN: Efficient Primitives for Deep Learning
Chetlur, S., Woolley, C., Vandermersch, P., Cohen, J., Tran, J. J., Catanzaro, B., & Shelhamer, E. (2014) · arXiv (Cornell University) · doi:10.48550/arxiv.1410.0759 · Cited by 943 · cited 2×
“The inputs can be off-loaded from the CPU or GPU to DRAM and processed by the accelerator.”
BackgroundShiDianNao: shifting vision processing closer to the sensor
Du, Z., Fasthuber, R., Chen, T., Ienne, P., Li, L., Luo, T., Feng, X., Chen, Y., & Temam, O. (2015) · Proceedings of the 42nd Annual International Symposium on Computer Architecture · doi:10.1145/2749469.2750389 · Cited by 933 · cited 1×
“However, due to differences in technology, hardware resources and system setup, a direct comparison between different implementations does not provide much insight into the relative energy efficiency of different dataflows.”
Self-citationGarp: a MIPS processor with a reconfigurable coprocessor
Hauser, J. R., & Wawrzynek, J. (2002) · Proceedings. The 5th Annual IEEE Symposium on Field-Programmable Custom Computing Machines Cat. No.97TB100186) · doi:10.1109/fpga.1997.624600 · Cited by 792 · cited 1×
“SAs come in two flavors: coarse-grained SAs that consist of tiled arrays of ALU-style PEs connected together via on-chip networks [27] [28] [29], and fine-grained SAs that are usually in the form of an FPGA.”
Background14.5 Eyeriss: An energy-efficient reconfigurable accelerator for deep convolutional neural networks
Chen, Y.-H., Krishna, T., Emer, J., & Sze, V. (2016) · 2016 IEEE International Solid-State Circuits Conference (ISSCC) · doi:10.1109/isscc.2016.7418007 · Cited by 713
Self-citationADRES: An Architecture with Tightly Coupled VLIW Processor and Coarse-Grained Reconfigurable Matrix
Mei, B., Vernalde, S., Verkest, D., De Man, H., & Lauwereins, R. (2003) · Lecture Notes in Computer Science · doi:10.1007/978-3-540-45234-8_7 · Cited by 558 · cited 1×
“SAs come in two flavors: coarse-grained SAs that consist of tiled arrays of ALU-style PEs connected together via on-chip networks [27] [28] [29], and fine-grained SAs that are usually in the form of an FPGA.”
BackgroundHandwritten digit recognition: applications of neural network chips and automatic learning
Le Cun, Y., Jackel, L. D., Boser, B., Denker, J. S., Graf, H. P., Guyon, I., Henderson, D., Howard, R. E., & Hubbard, W. (1989) · IEEE Communications Magazine · doi:10.1109/35.41400 · Cited by 480 · cited 1×
“These state-of-the-art CNNs [2– 5] are orders of magnitude larger than those used in the 1990s [9], requiring up to hundreds of megabytes for filter weight storage and 30k-600k operations per input pixel.”
BackgroundUnderstanding sources of inefficiency in general-purpose chips
Hameed, R., Qadeer, W., Wachs, M., Azizi, O., Solomatnikov, A., Lee, B. C., Richardson, S., Kozyrakis, C., & Horowitz, M. (2010) · Proceedings of the 37th annual international symposium on Computer architecture · doi:10.1145/1815961.1815968 · Cited by 391
Minimizing Computation in Convolutional Neural Networks
Cong, J., & Xiao, B. (2014) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-11179-7_36 · Cited by 372 · cited 1×
“Convolutions account for over 90% of the CNN operations and dominates runtime [10].”
BackgroundScaling to the end of silicon with EDGE architectures
Burger, D., Keckler, S. W., McKinley, K. S., Dahlin, M., John, L. K., Lin, C., Moore, C. R., Burrill, J., McDonald, R. G., & Yoder, W. (2004) · Computer · doi:10.1109/mc.2004.65 · Cited by 354 · cited 1×
“The expected performance advantage and large design space of coarse-grained SAs has inspired much research on the evaluation of its architectures, control schemes, operation scheduling and dataflow models [30] [31] [32] [33] [34] [35].”
BackgroundA dynamically configurable coprocessor for convolutional neural networks
Chakradhar, S., Sankaradas, M., Jakkula, V., & Cadambi, S. (2010) · Proceedings of the 37th annual international symposium on Computer architecture · doi:10.1145/1815961.1815993 · Cited by 343 · cited 1×
“Many previous papers have proposed specialized CNN dataflows on various platforms, including GPU [14], FPGA [15] [16] [17] [18] [19] [20] [21], and ASIC [22] [23] [24] [25] [26].”
BackgroundMemory-centric accelerator design for Convolutional Neural Networks
Peemen, M., Setio, A. A. A., Mesman, B., & Corporaal, H. (2013) · 2013 IEEE 31st International Conference on Computer Design (ICCD) · doi:10.1109/iccd.2013.6657019 · Cited by 332 · cited 1×
“Many previous papers have proposed specialized CNN dataflows on various platforms, including GPU [14], FPGA [15] [16] [17] [18] [19] [20] [21], and ASIC [22] [23] [24] [25] [26].”
BackgroundA 240 G-ops/s Mobile Coprocessor for Deep Neural Networks
Gokhale, V., Jin, J., Dundar, A., Martini, B., & Culurciello, E. (2014) · 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops · doi:10.1109/cvprw.2014.106 · Cited by 300 · cited 1×
“Many previous papers have proposed specialized CNN dataflows on various platforms, including GPU [14], FPGA [15] [16] [17] [18] [19] [20] [21], and ASIC [22] [23] [24] [25] [26].”
BackgroundMirsky, & DeHon (1996) · Proceedings IEEE Symposium on FPGAs for Custom Computing Machines FPGA-96 · doi:10.1109/fpga.1996.564808 · Cited by 281 · cited 1×
“SAs come in two flavors: coarse-grained SAs that consist of tiled arrays of ALU-style PEs connected together via on-chip networks [27] [28] [29], and fine-grained SAs that are usually in the form of an FPGA.”
BackgroundA Massively Parallel Coprocessor for Convolutional Neural Networks
Sankaradas, M., Jakkula, V., Cadambi, S., Chakradhar, S., Durdanovic, I., Cosatto, E., & Graf, H. P. (2009) · 2009 20th IEEE International Conference on Application-specific Systems, Architectures and Processors · doi:10.1109/asap.2009.25 · Cited by 224 · cited 3×
“FPGAs are also used to build CNN accelerators, and these designs usually use integrated DSP slices to construct the PE datapaths [15–21].”
BackgroundDynamically Specialized Datapaths for energy efficient computing
Govindaraju, V., Ho, C.-H., & Sankaralingam, K. (2011) · 2011 IEEE 17th International Symposium on High Performance Computer Architecture · doi:10.1109/hpca.2011.5749755 · Cited by 205 · cited 1×
“The expected performance advantage and large design space of coarse-grained SAs has inspired much research on the evaluation of its architectures, control schemes, operation scheduling and dataflow models [30] [31] [32] [33] [34] [35].”
BackgroundTowards energy-proportional datacenter memory with mobile DRAM
Malladi, K. T., Lee, B. C., Nothaft, F. A., Kozyrakis, C., Periyathambi, K., & Horowitz, M. (2012) · ACM SIGARCH Computer Architecture News · doi:10.1145/2366231.2337164 · Cited by 203 · cited 1×
“Similar to the energy consumption quantification in previous experiments [11, 12, 43], Table IV shows the energy cost of accessing each level relative to a MAC operation under the listed conditions.”
ResultConvolution engine: balancing efficiency & flexibility in specialized computing
Qadeer, W., Hameed, R., Shacham, O., Venkatesan, P., Kozyrakis, C., & Horowitz, M. A. (2013) · Proceedings of the 40th Annual International Symposium on Computer Architecture · doi:10.1145/2485922.2485925 · Cited by 168 · cited 1×
“Before CNNs became mainstream, there was already research on high-efficiency convolution due to its wide applicability in image signal processing (ISP) [40].”
BackgroundOrigami: A Convolutional Network Accelerator
Cavigelli, L., Gschwend, D., Mayer, C., Willi, S., Muheim, B., & Benini, L. (2015) · Proceedings of the 25th edition on Great Lakes Symposium on VLSI · doi:10.1145/2742060.2743766 · Cited by 156 · cited 1×
“Many previous papers have proposed specialized CNN dataflows on various platforms, including GPU [14], FPGA [15] [16] [17] [18] [19] [20] [21], and ASIC [22] [23] [24] [25] [26].”
BackgroundTowards energy-proportional datacenter memory with mobile DRAM
Malladi, K. T., Nothaft, F. A., Periyathambi, K., Lee, B. C., Kozyrakis, C., & Horowitz, M. (2012) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1109/isca.2012.6237004 · Cited by 156
PipeRench: A virtualized programmable datapath in 0.18 micron technology
Schmit, H. H., Whelihan, D. J., Tsai, A., Moe, M., Levine, B., & Taylor, R. R. (2003) · Custom Integrated Circuits Conference · doi:10.1109/cicc.2002.1012767 · Cited by 151 · cited 1×
“The expected performance advantage and large design space of coarse-grained SAs has inspired much research on the evaluation of its architectures, control schemes, operation scheduling and dataflow models [30] [31] [32] [33] [34] [35].”
BackgroundSwanson, S., Schwerin, A., Mercaldi, M., Petersen, A., Putnam, A. R., Michelson, K., Oskin, M., & Eggers, S. J. (2007) · ACM Transactions on Computer Systems · doi:10.1145/1233307.1233308 · Cited by 150 · cited 1×
“The expected performance advantage and large design space of coarse-grained SAs has inspired much research on the evaluation of its architectures, control schemes, operation scheduling and dataflow models [30] [31] [32] [33] [34] [35].”
BackgroundTriggered instructions: a control paradigm for spatially-programmed architectures
Parashar, A., Pellauer, M., Adler, M., Ahsan, B., Crago, N. C., Lustig, D., Pavlov, V., Zhai, A., Gambhir, M., Jaleel, A., et al. (2013) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1145/2485922.2485935 · Cited by 120 · cited 1×
“The expected performance advantage and large design space of coarse-grained SAs has inspired much research on the evaluation of its architectures, control schemes, operation scheduling and dataflow models [30] [31] [32] [33] [34] [35].”
BackgroundSelf-citationPark, S., Bong, K., Shin, D., Lee, J., Choi, S., & Yoo, H.-J. (2015) · 2015 IEEE International Solid-State Circuits Conference - (ISSCC) Digest of Technical Papers · doi:10.1109/isscc.2015.7062935 · Cited by 108
Power, Programmability, and Granularity: The Challenges of ExaScale Computing
Dally, B. (2011) · IEEE International Parallel and Distributed Processing Symposium (IPDPS) · doi:10.1109/ipdps.2011.420 · Cited by 62 · cited 1×
“Although these operations can leverage highly-parallel compute paradigms, such as SIMD/SIMT, throughput may not scale accordingly due to the accompanying bandwidth requirement, and the energy consumption remains high as data movement can be more expensive than computation [11] [12] [13].”
BackgroundExploring the potential of heterogeneous von neumann/dataflow execution models
Nowatzki, T., Gangadhar, V., & Sankaralingam, K. (2015) · Proceedings - International Symposium on Computer Architecture/Proceedings · doi:10.1145/2749469.2750380 · Cited by 60 · cited 1×
“The expected performance advantage and large design space of coarse-grained SAs has inspired much research on the evaluation of its architectures, control schemes, operation scheduling and dataflow models [30] [31] [32] [33] [34] [35].”
BackgroundTowards an embedded biologically-inspired machine vision processor
Sriram, V., Cox, D., Tsoi, K. H., & Luk, W. (2010) · 2010 International Conference on Field-Programmable Technology · doi:10.1109/fpt.2010.5681487 · Cited by 36 · cited 1×
“Many previous papers have proposed specialized CNN dataflows on various platforms, including GPU [14], FPGA [15] [16] [17] [18] [19] [20] [21], and ASIC [22] [23] [24] [25] [26].”
BackgroundExploiting spatial architectures for edit distance algorithms
Tithi, J. J., Crago, N. C., & Emer, J. (2014) · IEEE International Symposium on Performance Analysis of Systems and Software (ISPASS) · doi:10.1109/ispass.2014.6844458 · Cited by 17 · cited 1×
“The implementation of the RS dataflow in Eyeriss is inspired by the idea of applying a strip mining technique in a spatial architecture [42].”
MethodSelf-citation
Research works printed but not linked: 1
- S. Han H. Mao and W. J. Dally "Deep Compression: Compressing Deep Neural Network with Pruning Trained Quantization and Huffman Coding " in ICLR 2016. S. Han H. Mao and W. J. Dally "Deep Compression: Compressing Deep Neural Network with Pruning Trained Quantization and Huffman Coding " in ICLR 2016.
Cited by
In-Datacenter Performance Analysis of a Tensor Processing Unit
Jouppi, N. P., Young, C., Patil, N., Patterson, D., Agrawal, G., Bajwa, R., Bates, S., Bhatia, S., Boden, N., Borchers, A., et al. (2017) · Proceedings of the 44th Annual International Symposium on Computer Architecture · doi:10.1145/3079856.3080246 · Cited by 4,537
“The more recent DianNao family of NN architectures minimizes memory accesses both on the chip and to external DRAM by having efficient architectural support for the memory access patterns that appear in NN applications [28, 11].”
BackgroundEfficient Processing of Deep Neural Networks: A Tutorial and Survey
Sze, V., Chen, Y.-H., Yang, T.-J., & Emer, J. S. (2017) · Proceedings of the IEEE · doi:10.1109/jproc.2017.2761740 · Cited by 4,215
Self-citationEyeriss: An Energy-Efficient Reconfigurable Accelerator for Deep Convolutional Neural Networks
Chen, Y.-H., Krishna, T., Emer, J. S., & Sze, V. (2017) · IEEE Journal of Solid-State Circuits · doi:10.1109/jssc.2016.2616357 · Cited by 3,247
“It is reconfigurable for different shapes and optimizes for the best energy efficiency [32].”
BackgroundMethodInfluentialSelf-citationAn Introduction to Deep Learning for the Physical Layer
O'Shea, T., & Hoydis, J. (2017) · IEEE Transactions on Cognitive Communications and Networking · doi:10.1109/tccn.2017.2758370 · Cited by 2,980
Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing
Zhou, Z., Chen, X., Li, E., Zeng, L., Luo, K., & Zhang, J. (2019) · Proceedings of the IEEE · doi:10.1109/jproc.2019.2918951 · Cited by 2,455
“However, for energy-constrained end devices, the above magnitude-based weight pruning method may not be directly applicable since empirical measurements show that the reduction in the number of weights does not necessarily translate into significant energy saving [72].”
Background
Metadata
- Authors
- Chen, Y.-H., Emer, J., & Sze, V.
- Affiliations
- Massachusetts Institute of Technology (US)
- NVIDIA Research, NVIDIA
- Venue
- ACM SIGARCH Computer Architecture News
- Published
- 2016-06-18
- Volume
- 44
- Issue
- 3
- Pages
- 367-379
- ISSN
- 0163-5964 (ISSN-L)
- OpenAlex
- W2442974303
- Cited by
- 1,182
- References
- 48
- Citations by year
- 2026: 35 · 2025: 65 · 2024: 88 · 2023: 102 · 2022: 126 · 2021: 195 · 2020: 195 · 2019: 166 · 2018: 131 · 2017: 72 · 2016: 7
- FWCI
- 189.92
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- ACM SIGARCH Computer Architecture News (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Hardware and Architecture › Parallel Computing and Optimization Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Hardware and Architecture › Embedded Systems Design Techniques
- Keywords
- spatial parallelism, AlexNet
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
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.