DOCSAID Platform — visual assets and notices Brand mark and product icons Original SVG implementation for DOCSAID, based on the owner's previous blue/orange document brand. Source: frontend/src/Brand.tsx and favicon.svg. No claim of registered trademark status or exclusive rights to common shapes. Functional UI icons: Lucide 1.62.0 (ISC), including Feather-derived icons (MIT). Full notices: /licenses/lucide.txt Mathematical notation: KaTeX (MIT). Full notice: /licenses/katex.txt Research-note diagrams: Mermaid 12.2.0 (MIT), loaded only for Mermaid diagrams. Full notice: /licenses/mermaid.txt Sample document images: MIDV-2020 (CC BY-SA 2.5). Attribution and license links: /licenses/midv-2020.txt Synthetic faces in the dataset: https://generated.photos/ DM Sans and Manrope are self-hosted Google Fonts WOFF2 subsets, with system-font fallback. Their upstream licenses are SIL Open Font License 1.1. Full notices: /licenses/dmsans-OFL.txt and /licenses/manrope-OFL.txt Pinned sources and file SHA-256: docs/font-assets.json in the source repository. https://github.com/google/fonts/blob/main/ofl/dmsans/OFL.txt https://github.com/google/fonts/blob/main/ofl/manrope/OFL.txt Migrated research illustrations retain original captions and source links. They are not presented as newly created DOCSAID artwork. Their source and original-file checksums are preserved in archive/manifest.json. Blog paper-note illustrations were migrated from the owner's previous site. Original paper citations and existing captions remain with each article. Source paths and file checksums are recorded in content/paper-records/import-manifest.json in the repository. Article proofreading does not grant new rights to third-party figures; the open-model Apache licenses below do not apply to these illustrations. Unmodified public ONNX weights: DocsaidLab DocAligner and MRZScanner. Licensed under Apache License 2.0 by the project owner. Full licenses: /licenses/docaligner-Apache-2.0.txt and /licenses/mrzscanner-Apache-2.0.txt Pinned revisions, source URLs and SHA-256: /api/v1/catalog Training datasets are not included and retain their original providers' terms. Paper universe: Three.js 0.186.1 (MIT), Copyright Three.js authors. Full notice: /licenses/three.txt Procedural planet artwork, GLSL surfaces and SVG research-desk illustrations are generated by this repository. No third-party planet textures are included. Long-context article figures (2026-10-26): Position Interpolation, Figure 1: Shouyuan Chen, Sherman Wong, Liangjian Chen, Yuandong Tian (2023). Original unmodified figure from arXiv 2306.15595v2. https://arxiv.org/html/2306.15595v2 Lost in the Middle, Figure 1: Nelson F. Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, Percy Liang (2024). TACL 12:157–173, Association for Computational Linguistics. https://aclanthology.org/2024.tacl-1.9/ Rasterized and cropped to the complete plot, axes, title and legend; surrounding article text and English caption removed. No data redrawn. Both figures licensed under Creative Commons Attribution 4.0 International: https://creativecommons.org/licenses/by/4.0/ Source URLs, attribution, extraction details and SHA-256 are preserved in content/blog/sources/llm-long-context/figure-sources.json in the repository. DETR evolution article figure (2026-10-27): RF-DETR: Neural Architecture Search for Real-Time Detection Transformers Isaac Robinson, Peter Robicheaux, Matvei Popov, Deva Ramanan, Neehar Peri. arXiv:2511.09554v2, published at ICLR 2026. Source: https://arxiv.org/abs/2511.09554v2 License: Creative Commons Attribution 4.0 International https://creativecommons.org/licenses/by/4.0/ Figure 1, top-left COCO AP50:95 vs latency panel only. Cropped without relabeling or altering data; original legend retained. The original figure retains its authors' copyright. Source/crop/hash record: content/blog/sources/detr-evolution/figure-sources.json Long-context and DETR blog covers (2026-10-27): Original DOCSAID React/SVG article artwork adapted into language-neutral 8:5 index covers. Licensed under the repository's Apache-2.0 license. The existing Manrope font is used under SIL Open Font License 1.1; see manrope-OFL.txt. SVG sources, component/font hashes, dimensions and WebP hashes are recorded beside each manuscript in cover-source.svg and cover-source.json. Third-party paper figures retain their separate notices. RAG article and eleven new paper notes (2026-11-09): The RAG cover, four teaching SVGs and the NoLiMa numerical chart are original DOCSAID artwork under Apache-2.0, including localized label variants. NoLiMa uses only Table 3 scalar values; no NoLiMa paper image is reproduced. System Noto Sans CJK is used during chart rendering; font files are not bundled. The cover uses Manrope under the SIL OFL notice above. The following paper figures and tables are rasterized/cropped excerpts, with their original labels and data retained. Surrounding captions are translated. CC-licensed excerpts retain the authors’ licenses listed below; other excerpts are limited scholarly quotations accompanying discussion of the cited work. The arXiv distribution license is not treated as a general reuse license. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen-tau Yih, Tim Rocktäschel, Sebastian Riedel, Douwe Kiela https://arxiv.org/pdf/2005.11401v4 Excerpts: Figure 1, Figure 2, Table 1. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/rag/source-record.json ReAct: Synergizing Reasoning and Acting in Language Models Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, Yuan Cao https://arxiv.org/pdf/2210.03629v3 Excerpts: Figure 1 (part 1). License: https://creativecommons.org/licenses/by/4.0/ Provenance: content/paper-records/authored-sources/react/source-record.json Lost in the Middle: How Language Models Use Long Contexts Nelson F. Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, Percy Liang https://arxiv.org/pdf/2307.03172v3 Excerpts: Figure 5, Figure 7, Figure 11. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/lost-in-the-middle/source-record.json Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection Akari Asai, Zeqiu Wu, Yizhong Wang, Avirup Sil, Hannaneh Hajishirzi https://arxiv.org/pdf/2310.11511v1 Excerpts: Figure 1, Figure 3. License: https://creativecommons.org/licenses/by/4.0/ Provenance: content/paper-records/authored-sources/self-rag/source-record.json Seven Failure Points When Engineering a Retrieval Augmented Generation System Scott Barnett, Stefanus Kurniawan, Srikanth Thudumu, Zach Brannelly, Mohamed Abdelrazek https://arxiv.org/pdf/2401.05856v1 Excerpts: Figure 1. License: https://creativecommons.org/licenses/by/4.0/ Provenance: content/paper-records/authored-sources/rag-failure-points/source-record.json Retrieval Augmented Generation or Long-Context LLMs? A Comprehensive Study and Hybrid Approach Zhuowan Li, Cheng Li, Mingyang Zhang, Qiaozhu Mei, Michael Bendersky https://aclanthology.org/2024.emnlp-industry.66.pdf Excerpts: Table 1, Figure 2, Figure 4. License: https://creativecommons.org/licenses/by/4.0/ Provenance: content/paper-records/authored-sources/self-route/source-record.json Don’t Do RAG: When Cache-Augmented Generation is All You Need for Knowledge Tasks Brian J Chan, Chao-Ting Chen, Jui-Hung Cheng, Hen-Hsen Huang https://arxiv.org/pdf/2412.15605v2 Excerpts: Figure 1, Table 2, Table 3. License: https://creativecommons.org/licenses/by-sa/4.0/ Provenance: content/paper-records/authored-sources/cag/source-record.json Is Grep All You Need? How Agent Harnesses Reshape Agentic Search Sahil Sen, Akhil Kasturi, Elias Lumer, Anmol Gulati, Vamse Kumar Subbiah https://arxiv.org/pdf/2605.15184v1 Excerpts: Table 1. License: https://creativecommons.org/licenses/by/4.0/ Provenance: content/paper-records/authored-sources/is-grep-all-you-need/source-record.json BM25 Wins at Scale: A Scaling Study of Retrieval-Augmented Generation Paradigms Pengyu Wang, Benfeng Xu, Shaohan Wang, Mingxuan Du, Xin Zeng, Huarui Wu, Lei Zhang, Licheng Zhang https://arxiv.org/pdf/2607.26497v3 Excerpts: Figure 3. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/bm25-wins-at-scale/source-record.json Deep Agentic Search for Repository-Level Code Question Answering: An Empirical Study Amirkia Rafiei Oskooei, Bora Ilci, Alperen Kayim, Mehmet Egemen Uzun, Berat Can, Kaan Emre Kara, Ozan Orhan, Mehmet S. Aktas https://arxiv.org/pdf/2608.01507v1 Excerpts: Fig. 1, Fig. 19. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/deep-agentic-search/source-record.json CJK Markdown emphasis: remark-cjk-friendly 2.3.4 (MIT). Pinned dependency notices: /licenses/markdown-cjk.txt Source: https://github.com/tats-u/markdown-cjk-friendly Graph convolution tour and companion notes: The tour cover and 21 teaching SVGs, the six GCN primer SVGs, the three interactive GCN labs and the SAN diagram (including localized label variants) are original DOCSAID works under Apache-2.0. System fonts are used for rendering; no font file is embedded. SAN (CC BY-NC-SA 4.0) contributes concepts only; no paper image is reused. The following excerpts retain their original labels, with translated captions alongside commentary on each paper. Non-CC excerpts are limited scholarly quotations, not works under the site license. Predict then Propagate: Graph Neural Networks meet Personalized PageRank Johannes Gasteiger, Aleksandar Bojchevski, Stephan Günnemann https://arxiv.org/pdf/1810.05997v6 Excerpts: Figure 1, Figure 4. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/appnp/source-record.json Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering Michaël Defferrard, Xavier Bresson, Pierre Vandergheynst https://arxiv.org/pdf/1606.09375v3 Excerpts: Figure 1. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/chebnet/source-record.json Classic GNNs are Strong Baselines: Reassessing GNNs for Node Classification Yuankai Luo, Lei Shi, Xiao-Ming Wu https://arxiv.org/pdf/2406.08993v2 Excerpts: Figure 1. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/classic-gnns/source-record.json Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learning Qimai Li, Zhichao Han, Xiao-Ming Wu https://arxiv.org/pdf/1801.07606v1 Excerpts: Figure 1, Figure 2. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/deeper-insights/source-record.json Graph Attention Networks Petar Veličković, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Liò, Yoshua Bengio https://arxiv.org/pdf/1710.10903v3 Excerpts: Figure 1. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/gat/source-record.json Semi-Supervised Classification with Graph Convolutional Networks Thomas N. Kipf, Max Welling https://arxiv.org/pdf/1609.02907v4 Excerpts: Figure 1, Figure 5. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/gcn/source-record.json How Powerful are Graph Neural Networks? Keyulu Xu, Weihua Hu, Jure Leskovec, Stefanie Jegelka https://arxiv.org/pdf/1810.00826v3 Excerpts: Figure 2, Figure 3. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/gin/source-record.json Can Classic GNNs Be Strong Baselines for Graph-level Tasks? Simple Architectures Meet Excellence Yuankai Luo, Lei Shi, Xiao-Ming Wu https://arxiv.org/pdf/2502.09263v3 Excerpts: Figure 2. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/gnn-plus/source-record.json A Generalization of Transformer Networks to Graphs Vijay Prakash Dwivedi, Xavier Bresson https://arxiv.org/pdf/2012.09699v2 Excerpts: Figure 1. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/graph-transformer/source-record.json Recipe for a General, Powerful, Scalable Graph Transformer Ladislav Rampášek, Mikhail Galkin, Vijay Prakash Dwivedi, Anh Tuan Luu, Guy Wolf, Dominique Beaini https://arxiv.org/pdf/2205.12454v4 Excerpts: Figure 1. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/graphgps/source-record.json Do Transformers Really Perform Bad for Graph Representation? Chengxuan Ying, Tianle Cai, Shengjie Luo, Shuxin Zheng, Guolin Ke, Di He, Yanming Shen, Tie-Yan Liu https://arxiv.org/pdf/2106.05234v5 Excerpts: Figure 1, Figure 2. License: https://creativecommons.org/licenses/by-sa/4.0/ Provenance: content/paper-records/authored-sources/graphormer/source-record.json Inductive Representation Learning on Large Graphs William L. Hamilton, Rex Ying, Jure Leskovec https://arxiv.org/pdf/1706.02216v4 Excerpts: Figure 1, Figure 2. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/graphsage/source-record.json Long Range Graph Benchmark Vijay Prakash Dwivedi, Ladislav Rampášek, Mikhail Galkin, Ali Parviz, Guy Wolf, Anh Tuan Luu, Dominique Beaini https://arxiv.org/pdf/2206.08164v4 Excerpts: Figure 2. License: https://creativecommons.org/licenses/by/4.0/ Provenance: content/paper-records/authored-sources/lrgb/source-record.json Neural Message Passing for Quantum Chemistry Justin Gilmer, Samuel S. Schoenholz, Patrick F. Riley, Oriol Vinyals, George E. Dahl https://arxiv.org/pdf/1704.01212v2 Excerpts: Figure 1. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/mpnn/source-record.json On the Bottleneck of Graph Neural Networks and its Practical Implications Uri Alon, Eran Yahav https://arxiv.org/pdf/2006.05205v4 Excerpts: Figure 1, Figure 3. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/over-squashing/source-record.json Simplifying Graph Convolutional Networks Felix Wu, Tianyi Zhang, Amauri Holanda de Souza, Christopher Fifty, Tao Yu, Kilian Q. Weinberger https://arxiv.org/pdf/1902.07153v2 Excerpts: Figure 1, Figure 2. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/sgc/source-record.json Spectral Networks and Locally Connected Networks on Graphs Joan Bruna, Wojciech Zaremba, Arthur Szlam, Yann LeCun https://arxiv.org/pdf/1312.6203v3 Excerpts: Figure 2, Figure 6. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/spectral-networks/source-record.json Spatial Temporal Graph Convolutional Networks for Skeleton-Based Action Recognition Sijie Yan, Yuanjun Xiong, Dahua Lin https://arxiv.org/pdf/1801.07455v2 Excerpts: Figure 1, Figure 3. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/st-gcn/source-record.json Where Did the Gap Go? Reassessing the Long-Range Graph Benchmark Jan Tönshoff, Martin Ritzert, Eran Rosenbluth, Martin Grohe https://arxiv.org/pdf/2309.00367v2 Excerpts: Figure 1. License: https://creativecommons.org/licenses/by/4.0/ Provenance: content/paper-records/authored-sources/where-did-the-gap-go/source-record.json Classic CNN notes, batches 5 and 6 InceptionNeXt: When Inception Meets ConvNeXt Weihao Yu, Pan Zhou, Shuicheng Yan, Xinchao Wang https://arxiv.org/pdf/2303.16900v3 Excerpts: Figure 1, Figure 2. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/inceptionnext/source-record.json ConvNets Match Vision Transformers at Scale Samuel L. Smith, Andrew Brock, Leonard Berrada, Soham De https://arxiv.org/pdf/2310.16764v1 Excerpts: Figure 1, Figure 2, Figure 3. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/convnets-match-vits/source-record.json Efficient Deformable ConvNets: Rethinking Dynamic and Sparse Operator for Vision Applications Yuwen Xiong, Zhiqi Li, Yuntao Chen, Feng Wang, Xizhou Zhu, Jiapeng Luo, Wenhai Wang, Tong Lu, Hongsheng Li, Yu Qiao, Lewei Lu, Jie Zhou, Jifeng Dai https://arxiv.org/pdf/2401.06197v1 Excerpts: Figure 1, Figure 2, Figure 3. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/dcnv4/source-record.json DenseNets Reloaded: Paradigm Shift Beyond ResNets and ViTs Donghyun Kim, Byeongho Heo, Dongyoon Han https://arxiv.org/pdf/2403.19588v2 Excerpts: Fig. 1. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/rdnet/source-record.json OverLoCK: An Overview-first-Look-Closely-next ConvNet with Context-Mixing Dynamic Kernels Meng Lou, Yizhou Yu https://arxiv.org/pdf/2502.20087v3 Excerpts: Figure 3, Figure 5. License: https://creativecommons.org/licenses/by/4.0/ Provenance: content/paper-records/authored-sources/overlock/source-record.json Reviving ConvNeXt for Efficient Convolutional Diffusion Models Taesung Kwon, Lorenzo Bianchi, Lennart Wittke, Felix Watine, Fabio Carrara, Jong Chul Ye, Romann Weber, Vinicius Azevedo https://arxiv.org/pdf/2603.09408v1 Illustration: original DOCSAID FCDM block and U-shaped layout teaching figure; no paper figures reproduced. Original illustration license: Apache-2.0 (https://github.com/DocsaidLab/DocsaidPlatform/blob/main/LICENSE). Reproducible SVG and generator: content/paper-records/illustration-sources/fcdm/. Provenance: content/paper-records/authored-sources/fcdm/source-record.json WBMM: Windowed Batch Matrix Multiplication for Efficient Large Receptive Field Convolution Wan Song, Wei Zhou, Rui Wang, Jun Yu, Toru Kurihara, Jiajia Xu, Shu Zhan https://arxiv.org/pdf/2607.02097v2 Excerpts: Figure 1, Figure 2. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/wbmm/source-record.json Rewrite the Stars Xu Ma, Xiyang Dai, Yue Bai, Yizhou Wang, Yun Fu https://arxiv.org/pdf/2403.19967v1 Excerpts: Figure 1, Figure 2, Figure 3. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/starnet/source-record.json LSNet: See Large, Focus Small Ao Wang, Hui Chen, Zijia Lin, Jungong Han, Guiguang Ding https://arxiv.org/pdf/2503.23135v1 Excerpts: Figure 1, Figure 3. License: https://creativecommons.org/licenses/by/4.0/ Provenance: content/paper-records/authored-sources/lsnet/source-record.json UniRepLKNet: A Universal Perception Large-Kernel ConvNet for Audio, Video, Point Cloud, Time-Series and Image Recognition Xiaohan Ding, Yiyuan Zhang, Yixiao Ge, Sijie Zhao, Lin Song, Xiangyu Yue, Ying Shan https://arxiv.org/pdf/2311.15599v2 Excerpts: Figure 1, Figure 2. Copyright remains with the original authors; scholarly quotation only. Provenance: content/paper-records/authored-sources/unireplknet/source-record.json Blog original paper figure/table quotations These PNG excerpts retain their original authorship and are excluded from DOCSAID’s Apache-2.0 grant. llm-long-context/img/paper-ruler-results.png RULER: What's the Real Context Size of Your Long-Context Language Models?, Table 3 Hsieh, Cheng-Ping, Sun, Simeng, Kriman, Samuel, Acharya, Shantanu, Rekesh, Dima, Jia, Fei, Ginsburg, Boris https://arxiv.org/pdf/2404.06654v1#page=6 https://creativecommons.org/licenses/by/4.0/ detr-evolution/img/paper-detr-nms.png End-to-End Object Detection with Transformers, Figure 4 Carion, Nicolas, Massa, Francisco, Synnaeve, Gabriel, Usunier, Nicolas, Kirillov, Alexander, Zagoruyko, Sergey https://arxiv.org/pdf/2005.12872v3#page=12 https://creativecommons.org/publicdomain/zero/1.0/ detr-evolution/img/paper-hdetr-hybrid.png DETRs with Hybrid Matching, Figure 3 Jia, Ding, Yuan, Yuhui, He, Haodi, Wu, Xiaopei, Yu, Haojun, Lin, Weihong, Sun, Lei, Zhang, Chao, Hu, Han https://arxiv.org/pdf/2207.13080v3#page=4 https://creativecommons.org/licenses/by/4.0/ detr-evolution/img/paper-rtdetr-nms.png DETRs Beat YOLOs on Real-time Object Detection, Table 1 Zhao, Yian, Lv, Wenyu, Xu, Shangliang, Wei, Jinman, Wang, Guanzhong, Dang, Qingqing, Liu, Yi, Chen, Jie https://arxiv.org/pdf/2304.08069v3#page=3 Selected excerpt for scholarly commentary; no open redistribution licence claimed. detr-evolution/img/paper-yolov10-heads.png YOLOv10: Real-Time End-to-End Object Detection, Figure 2 Wang, Ao, Chen, Hui, Liu, Lihao, Chen, Kai, Lin, Zijia, Han, Jungong, Ding, Guiguang https://arxiv.org/pdf/2405.14458v2#page=4 https://creativecommons.org/licenses/by/4.0/ llm-rag-rise-and-decline/img/paper-selfroute-results.png Retrieval Augmented Generation or Long-Context LLMs? A Comprehensive Study and Hybrid Approach, Table 1 Zhuowan Li, Cheng Li, Mingyang Zhang, Qiaozhu Mei, Michael Bendersky https://aclanthology.org/2024.emnlp-industry.66.pdf#page=4 https://creativecommons.org/licenses/by/4.0/ ocr-language-prior/img/paper-vocabulary-examples.png On Vocabulary Reliance in Scene Text Recognition, Figure 1 Wan, Zhaoyi, Zhang, Jielei, Zhang, Liang, Luo, Jiebo, Yao, Cong https://arxiv.org/pdf/2005.03959v1#page=1 Selected excerpt for scholarly commentary; no open redistribution licence claimed. ocr-language-prior/img/paper-crutch-example.png Visual Merit or Linguistic Crutch? A Close Look at DeepSeek-OCR, Figure 1 Yunhao Liang, Ruixuan Ying, Bo Li, Hong Li, Kai Yan, Qingwen Li, Min Yang, Okamoto Satoshi, Zhe Cui, Shiwen Ni https://arxiv.org/pdf/2601.03714v2#page=1 Selected excerpt for scholarly commentary; no open redistribution licence claimed. ocr-language-prior/img/paper-faithc4-errors.png Do VLMs Read or Rewrite? On Transcription Faithfulness in Vision-Language Models, Table 5 Gwang Gook Lee, Kenan Emir Ak, Jay Mohta, Yan Xu, Dimitrios Dimitriadis https://arxiv.org/pdf/2607.21617v2#page=6 https://creativecommons.org/licenses/by/4.0/ where-did-mamba-go/img/paper-copy-phonebook.png Repeat After Me: Transformers are Better than State Space Models at Copying, Figure 1 Samy Jelassi, David Brandfonbrener, Sham M. Kakade, Eran Malach https://arxiv.org/pdf/2402.01032v2#page=2 https://creativecommons.org/licenses/by/4.0/ where-did-mamba-go/img/paper-mamba-format.png An Empirical Study of Mamba-based Language Models, Table 4 Waleffe, Roger, Byeon, Wonmin, Riach, Duncan, Norick, Brandon, Korthikanti, Vijay, Dao, Tri, Gu, Albert, Hatamizadeh, Ali, Singh, Sudhakar, Narayanan, Deepak, Kulshreshtha, Garvit, Singh, Vartika, Casper, Jared, Kautz, Jan, Shoeybi, Mohammad, Catanzaro, Bryan https://arxiv.org/pdf/2406.07887v1#page=7 Selected excerpt for scholarly commentary; no open redistribution licence claimed. where-did-mamba-go/img/paper-mamba3-retrieval.png Mamba-3: Improved Sequence Modeling using State Space Principles, Table 4 Aakash Lahoti, Kevin Y. Li, Berlin Chen, Caitlin Wang, Aviv Bick, J. Zico Kolter, Tri Dao, Albert Gu https://arxiv.org/pdf/2603.15569v1#page=14 https://creativecommons.org/licenses/by/4.0/ where-did-mamba-go/img/paper-mamba3-latency.png Mamba-3: Improved Sequence Modeling using State Space Principles, Table 7 Aakash Lahoti, Kevin Y. Li, Berlin Chen, Caitlin Wang, Aviv Bick, J. Zico Kolter, Tri Dao, Albert Gu https://arxiv.org/pdf/2603.15569v1#page=15 https://creativecommons.org/licenses/by/4.0/ graph-convolution-tour/img/paper-oversquash-results.png On the Bottleneck of Graph Neural Networks and its Practical Implications, Figure 3 Alon, Uri, Yahav, Eran https://arxiv.org/pdf/2006.05205v4#page=5 Selected excerpt for scholarly commentary; no open redistribution licence claimed. graph-convolution-tour/img/paper-graphormer-ablation.png Do Transformers Really Perform Bad for Graph Representation?, Table 5 Ying, Chengxuan, Cai, Tianle, Luo, Shengjie, Zheng, Shuxin, Ke, Guolin, He, Di, Shen, Yanming, Liu, Tie-Yan https://arxiv.org/pdf/2106.05234v5#page=9 https://creativecommons.org/licenses/by-sa/4.0/ graph-convolution-tour/img/paper-lrgb-reassessment.png Where Did the Gap Go? Reassessing the Long-Range Graph Benchmark, Figure 1 Tönshoff, Jan, Ritzert, Martin, Rosenbluth, Eran, Grohe, Martin https://arxiv.org/pdf/2309.00367v2#page=2 https://creativecommons.org/licenses/by/4.0/ fas-paper-roadmap/img/paper-cefa-modalities.png CASIA-SURF CeFA: A Benchmark for Multi-modal Cross-ethnicity Face Anti-spoofing, Figure 1 Li, Ajian, Tan, Zichang, Li, Xuan, Wan, Jun, Escalera, Sergio, Guo, Guodong, Li, Stan Z. https://arxiv.org/pdf/2003.05136v1#page=2 Selected excerpt for scholarly commentary; no open redistribution licence claimed. fas-paper-roadmap/img/paper-ssdg-domains.png Single-Side Domain Generalization for Face Anti-Spoofing, Figure 1 Jia, Yunpei, Zhang, Jie, Shan, Shiguang, Chen, Xilin https://arxiv.org/pdf/2004.14043v1#page=1 Selected excerpt for scholarly commentary; no open redistribution licence claimed. fas-paper-roadmap/img/paper-cdcn-difference.png Searching Central Difference Convolutional Networks for Face Anti-Spoofing, Figure 2 Yu, Zitong, Zhao, Chenxu, Wang, Zezheng, Qin, Yunxiao, Su, Zhuo, Li, Xiaobai, Zhou, Feng, Zhao, Guoying https://arxiv.org/pdf/2003.04092v1#page=3 Selected excerpt for scholarly commentary; no open redistribution licence claimed. RAG article also reuses Liu et al., Lost in the Middle, TACL 2024, Figure 1, CC BY 4.0 (https://aclanthology.org/2024.tacl-1.9/). The NoLiMa length chart is a DOCSAID original chart of Table 3 numerical facts, Apache-2.0; no NoLiMa figure pixels are reproduced. --- DOCSAID model-merging paper-figure notices --- Model merging article: original paper figure/table quotations The PNG quotations below retain the original authorship and are excluded from DOCSAID's Apache-2.0 grant. Timur Garipov, Pavel Izmailov, Dmitrii Podoprikhin, Dmitry Vetrov, Andrew Gordon Wilson, Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs, Figure 1 (arXiv 1802.10026v4) https://arxiv.org/pdf/1802.10026v4#page=2 Selected figure excerpt for scholarly commentary; no open redistribution licence claimed. Pavel Izmailov, Dmitrii Podoprikhin, Timur Garipov, Dmitry Vetrov, Andrew Gordon Wilson, Averaging Weights Leads to Wider Optima and Better Generalization, Figure 1 (arXiv 1803.05407v3) https://arxiv.org/pdf/1803.05407v3#page=2 Selected figure excerpt for scholarly commentary; no open redistribution licence claimed. Jonathan Frankle, Gintare Karolina Dziugaite, Daniel M. Roy, Michael Carbin, Linear Mode Connectivity and the Lottery Ticket Hypothesis, Figure 3 (arXiv 1912.05671v4) https://arxiv.org/pdf/1912.05671v4#page=4 Selected figure excerpt for scholarly commentary; no open redistribution licence claimed. Mitchell Wortsman, Gabriel Ilharco, Jong Wook Kim, Mike Li, Simon Kornblith, Rebecca Roelofs, Raphael Gontijo-Lopes, Hannaneh Hajishirzi, Ali Farhadi, Hongseok Namkoong, Ludwig Schmidt, Robust fine-tuning of zero-shot models, Figure 6 (arXiv 2109.01903v3) https://arxiv.org/pdf/2109.01903v3#page=11 Selected figure excerpt for scholarly commentary; no open redistribution licence claimed. Samuel K. Ainsworth, Jonathan Hayase, Siddhartha Srinivasa, Git Re-Basin: Merging Models modulo Permutation Symmetries, Figure 2 (arXiv 2209.04836v6) https://arxiv.org/pdf/2209.04836v6#page=5 CC BY 4.0: https://creativecommons.org/licenses/by/4.0/ Gabriel Ilharco, Marco Tulio Ribeiro, Mitchell Wortsman, Suchin Gururangan, Ludwig Schmidt, Hannaneh Hajishirzi, Ali Farhadi, Editing Models with Task Arithmetic, Figure 3 (arXiv 2212.04089v3) https://arxiv.org/pdf/2212.04089v3#page=5 CC BY 4.0: https://creativecommons.org/licenses/by/4.0/ Le Yu, Bowen Yu, Haiyang Yu, Fei Huang, Yongbin Li, Language Models are Super Mario: Absorbing Abilities from Homologous Models as a Free Lunch, Figure 1 (arXiv 2311.03099v3) https://arxiv.org/pdf/2311.03099v3#page=1 Selected figure excerpt for scholarly commentary; no open redistribution licence claimed. Takuya Akiba, Makoto Shing, Yujin Tang, Qi Sun, David Ha, Evolutionary Optimization of Model Merging Recipes, Figure 1 (arXiv 2403.13187v2) https://arxiv.org/pdf/2403.13187v2#page=5 Selected figure excerpt for scholarly commentary; no open redistribution licence claimed. Prateek Yadav, Tu Vu, Jonathan Lai, Alexandra Chronopoulou, Manaal Faruqui, Mohit Bansal, Tsendsuren Munkhdalai, What Matters for Model Merging at Scale?, Figure 7 (arXiv 2410.03617v1) https://arxiv.org/pdf/2410.03617v1#page=10 CC BY 4.0: https://creativecommons.org/licenses/by/4.0/ Oğuz Kağan Hitit, Leander Girrbach, Zeynep Akata, A Systematic Study of In-the-Wild Model Merging for Large Language Models, Figure 11 (arXiv 2511.21437v2) https://arxiv.org/pdf/2511.21437v2#page=26 Selected figure excerpt for scholarly commentary; no open redistribution licence claimed. Oğuz Kağan Hitit, Leander Girrbach, Zeynep Akata, A Systematic Study of In-the-Wild Model Merging for Large Language Models, Figure 12 (arXiv 2511.21437v2) https://arxiv.org/pdf/2511.21437v2#page=26 Selected figure excerpt for scholarly commentary; no open redistribution licence claimed. Mitchell Wortsman, Gabriel Ilharco, Samir Yitzhak Gadre, Rebecca Roelofs, Raphael Gontijo-Lopes, Ari S. Morcos, Hongseok Namkoong, Ali Farhadi, Yair Carmon, Simon Kornblith, Ludwig Schmidt, Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time, Table 3 (arXiv 2203.05482v3) https://arxiv.org/pdf/2203.05482v3#page=6 Selected figure excerpt for scholarly commentary; no open redistribution licence claimed. Prateek Yadav, Derek Tam, Leshem Choshen, Colin Raffel, Mohit Bansal, TIES-Merging: Resolving Interference When Merging Models, Table 1 (arXiv 2306.01708v2) https://arxiv.org/pdf/2306.01708v2#page=6 Selected figure excerpt for scholarly commentary; no open redistribution licence claimed. H. Brendan McMahan, Eider Moore, Daniel Ramage, Seth Hampson, Blaise Agüera y Arcas, Communication-Efficient Learning of Deep Networks from Decentralized Data, Figure 1 (arXiv 1602.05629v4) https://arxiv.org/pdf/1602.05629v4#page=4 Selected figure excerpt for scholarly commentary; no open redistribution licence claimed. Team Cohere, Command A: An Enterprise-Ready Large Language Model, Figure 12 (arXiv 2504.00698v2) https://arxiv.org/pdf/2504.00698v2#page=35 CC BY 4.0: https://creativecommons.org/licenses/by/4.0/ --- END DOCSAID model-merging paper-figure notices --- LinkedIn author-profile icon: Font Awesome Free 7.3.1 by Fonticons, Inc. Copyright 2026 Fonticons, Inc. SVG icon licensed under Creative Commons Attribution 4.0 International: https://creativecommons.org/licenses/by/4.0/ Original source: https://github.com/FortAwesome/Font-Awesome/blob/7.x/svgs/brands/linkedin.svg Fixed source blob: 387b36f26c5d6b4c9c4011da7214fdf01b121a5e Changes: crop vertical canvas whitespace and use LinkedIn blue (#0a66c2); the original icon path and copyright comment are retained. The complete upstream license is included in fontawesome.txt.