{"id":18852793,"url":"https://github.com/tech-srl/how_attentive_are_gats","last_synced_at":"2025-07-03T08:07:30.443Z","repository":{"id":41507600,"uuid":"370004769","full_name":"tech-srl/how_attentive_are_gats","owner":"tech-srl","description":"Code for the paper \"How Attentive are Graph Attention Networks?\" (ICLR'2022)","archived":false,"fork":false,"pushed_at":"2022-04-05T15:47:49.000Z","size":3500,"stargazers_count":335,"open_issues_count":4,"forks_count":41,"subscribers_count":11,"default_branch":"main","last_synced_at":"2025-05-18T06:05:29.421Z","etag":null,"topics":["are","attention","attentive","gat","gatv2","graph","graph-attention-networks","graph-neural-networks","how","networks","pytorch"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/tech-srl.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":"CITATION.cff","codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2021-05-23T08:59:00.000Z","updated_at":"2025-05-14T14:31:11.000Z","dependencies_parsed_at":"2022-08-31T21:21:07.649Z","dependency_job_id":null,"html_url":"https://github.com/tech-srl/how_attentive_are_gats","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tech-srl%2Fhow_attentive_are_gats","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tech-srl%2Fhow_attentive_are_gats/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tech-srl%2Fhow_attentive_are_gats/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/tech-srl%2Fhow_attentive_are_gats/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/tech-srl","download_url":"https://codeload.github.com/tech-srl/how_attentive_are_gats/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":256205923,"owners_count":22353267,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["are","attention","attentive","gat","gatv2","graph","graph-attention-networks","graph-neural-networks","how","networks","pytorch"],"created_at":"2024-11-08T03:41:34.410Z","updated_at":"2025-07-03T08:07:30.395Z","avatar_url":"https://github.com/tech-srl.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# How Attentive are Graph Attention Networks?\n\nThis repository is the official implementation of [How Attentive are Graph Attention Networks?](https://arxiv.org/pdf/2105.14491.pdf). \n\n**_January 2022_**: the paper was accepted to **ICLR'2022** !\n\n![alt text](images/fig1.png \"Figure 1 from the paper\")\n\n\n## Using GATv2\n\n**GATv2 is now available as part of PyTorch Geometric library!** \n```\nfrom torch_geometric.nn.conv.gatv2_conv import GATv2Conv\n```\n\n[https://pytorch-geometric.readthedocs.io/en/latest/modules/nn.html#torch_geometric.nn.conv.GATv2Conv](https://pytorch-geometric.readthedocs.io/en/latest/modules/nn.html#torch_geometric.nn.conv.GATv2Conv)\n\nand also is [in this main directory](gatv2_conv_PyG.py).\n\n**GATv2 is now available as part of DGL library!** \n```\nfrom dgl.nn.pytorch import GATv2Conv\n```\n\n[https://docs.dgl.ai/en/latest/api/python/nn.pytorch.html#gatv2conv](https://docs.dgl.ai/en/latest/api/python/nn.pytorch.html#gatv2conv)\n\nand also in [this repository](gatv2_conv_DGL.py).\n\n**GATv2 is now available as part of Google's TensorFlow GNN library!** \n```\nfrom tensorflow_gnn.graph.keras.layers.gat_v2 import GATv2Convolution\n```\n\n[https://github.com/tensorflow/gnn/blob/main/tensorflow_gnn/docs/api_docs/python/gnn/keras/layers/GATv2.md](https://github.com/tensorflow/gnn/blob/main/tensorflow_gnn/docs/api_docs/python/gnn/keras/layers/GATv2.md)\n\n## Code Structure\n\nSince our experiments (Section 4) are based on different frameworks, this repository is divided into several sub-projects:\n1. The subdirectory `arxiv_mag_products_collab_citation2_noise` contains the needed files to reproduce the results of \nNode-Prediction, Link-Prediction, and Robustness to Noise (Tables 2a, 3 and Figure 4).\n2. The subdirectory `proteins` contains the needed files to reproduce the results of ogbn-proteins in Node-Prediction (Table 2b).\n3. The subdirectory `dictionary_lookup` contains the need files to reproduce the results of the DictionaryLookup benchmark (Figure 3).\n4. The subdirectory `tf-gnn-samples` contains the needed files to reproduce the results of the VarMisuse and QM9 datasets \n(Table 1 and Table 4).\n\n## Requirements\nEach subdirectory contains its own requirements and dependencies.\n\nGenerally, all subdirectories depend on PyTorch 1.7.1 and [PyTorch Geometric](https://pytorch-geometric.readthedocs.io/) version 1.7.0 (`proteins` depends on [DGL](https://www.dgl.ai/) version 0.6.0).\nThe subdirectory `tf-gnn-samples` (VarMisuse and QM9) depends on TensorFlow 1.13. \n\n## Hardware\nIn general, all experiments can run on either GPU or CPU. \n\n\n## Citation\n[How Attentive are Graph Attention Networks?](https://arxiv.org/pdf/2105.14491.pdf)\n```\n@inproceedings{\n  brody2022how,\n  title={How Attentive are Graph Attention Networks? },\n  author={Shaked Brody and Uri Alon and Eran Yahav},\n  booktitle={International Conference on Learning Representations},\n  year={2022},\n  url={https://openreview.net/forum?id=F72ximsx7C1}\n}\n```\n\n\n\n\n\n\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftech-srl%2Fhow_attentive_are_gats","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ftech-srl%2Fhow_attentive_are_gats","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ftech-srl%2Fhow_attentive_are_gats/lists"}