{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/graph-inductive-biases-in-transformers","title":"Graph Inductive Biases in Transformers without Message Passing","arxiv_id":"2305.17589","date":"2023-05-27","proceeding":null,"authors":["Liheng Ma","Chen Lin","Derek Lim","Adriana Romero-Soriano","Puneet K. Dokania","Mark Coates","Philip Torr","Ser-Nam Lim"],"abstract":"Transformers for graph data are increasingly widely studied and successful in numerous learning tasks. Graph inductive biases are crucial for Graph Transformers, and previous works incorporate them using message-passing modules and/or positional encodings. However, Graph Transformers that use message-passing inherit known issues of message-passing, and differ significantly from Transformers used in other domains, thus making transfer of research advances more difficult. On the other hand, Graph Transformers without message-passing often perform poorly on smaller datasets, where inductive biases are more crucial. To bridge this gap, we propose the Graph Inductive bias Transformer (GRIT) -- a new Graph Transformer that incorporates graph inductive biases without using message passing. GRIT is based on several architectural changes that are each theoretically and empirically justified, including: learned relative positional encodings initialized with random walk probabilities, a flexible attention mechanism that updates node and node-pair representations, and injection of degree information in each layer. We prove that GRIT is expressive -- it can express shortest path distances and various graph propagation matrices. GRIT achieves state-of-the-art empirical performance across a variety of graph datasets, thus showing the power that Graph Transformers without message-passing can deliver.","url_abs":"https://arxiv.org/abs/2305.17589v1","url_pdf":"https://arxiv.org/pdf/2305.17589v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"graph-inductive-biases-in-transformers","repo_url":"https://github.com/liamma/grit","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"graph-inductive-biases-in-transformers","repo_url":"https://github.com/linusbao/MoSE","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"graph-classification","task_name":"Graph Classification"},{"task_slug":"graph-regression","task_name":"Graph Regression"},{"task_slug":"inductive-bias","task_name":"Inductive Bias"},{"task_slug":"node-classification","task_name":"Node Classification"}],"methods":[{"method_slug":"absolute-position-encodings","method_name":"Absolute Position Encodings"},{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"graph-transformer","method_name":"Graph Transformer"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"lapeigen","method_name":"LapEigen"},{"method_slug":"laplacian-pe","method_name":"Laplacian PE"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"position-wise-feed-forward-layer","method_name":"Position-Wise Feed-Forward Layer"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"transformer","method_name":"Transformer"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/graph-classification-on-cifar10-100k","task":"Graph Classification","dataset":"CIFAR10 100k","model":"GRIT","rank_in_archive_order":4,"of":20,"metrics":{"Accuracy (%)":"76.468"},"uses_additional_data":false},{"leaderboard":"/sota/graph-classification-on-mnist","task":"Graph Classification","dataset":"MNIST","model":"GRIT","rank_in_archive_order":11,"of":13,"metrics":{"Accuracy":"98.108"},"uses_additional_data":false},{"leaderboard":"/sota/graph-classification-on-peptides-func","task":"Graph Classification","dataset":"Peptides-func","model":"GRIT","rank_in_archive_order":14,"of":44,"metrics":{"AP":"0.6988±0.0082"},"uses_additional_data":false},{"leaderboard":"/sota/graph-regression-on-pcqm4mv2-lsc","task":"Graph Regression","dataset":"PCQM4Mv2-LSC","model":"GRIT","rank_in_archive_order":11,"of":20,"metrics":{"Validation MAE":"0.0859"},"uses_additional_data":false},{"leaderboard":"/sota/graph-regression-on-peptides-struct","task":"Graph Regression","dataset":"Peptides-struct","model":"GRIT","rank_in_archive_order":10,"of":39,"metrics":{"MAE":"0.2460±0.0012"},"uses_additional_data":false},{"leaderboard":"/sota/graph-regression-on-zinc","task":"Graph Regression","dataset":"ZINC","model":"GRIT","rank_in_archive_order":4,"of":27,"metrics":{"MAE":"0.059"},"uses_additional_data":false},{"leaderboard":"/sota/graph-regression-on-zinc-500k","task":"Graph Regression","dataset":"ZINC-500k","model":"GRIT","rank_in_archive_order":4,"of":36,"metrics":{"MAE":"0.059"},"uses_additional_data":false},{"leaderboard":"/sota/graph-regression-on-zinc-full","task":"Graph Regression","dataset":"ZINC-full","model":"GRIT","rank_in_archive_order":6,"of":19,"metrics":{"Test MAE":"0.023"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-cluster","task":"Node Classification","dataset":"CLUSTER","model":"GRIT","rank_in_archive_order":1,"of":12,"metrics":{"Accuracy":"80.026"},"uses_additional_data":false},{"leaderboard":"/sota/node-classification-on-pattern","task":"Node Classification","dataset":"PATTERN","model":"GRIT","rank_in_archive_order":2,"of":11,"metrics":{"Accuracy":"87.196"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2305.17589","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2305.17589"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/linusbao/MoSE","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/liamma/grit","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":6,"unverified":6},"by_repo_kind":{"official":{"samples":10,"ran":6,"repositories":1},"listed":{"samples":2,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"ecf14287d3465c48","entry":"accuracy_SBM","repo":"liamma/grit","repo_kind":"official","path":"grit/logger.py","file_url":"https://github.com/liamma/grit/blob/HEAD/grit/logger.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ecf14287d3465c48"}},{"code_sha256_prefix":"9dc78001d467fc9d","entry":"eval_spearmanr","repo":"liamma/grit","repo_kind":"official","path":"grit/logger.py","file_url":"https://github.com/liamma/grit/blob/HEAD/grit/logger.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9dc78001d467fc9d"}},{"code_sha256_prefix":"1bb331bf70a78136","entry":"get_final_pretrained_ckpt","repo":"liamma/grit","repo_kind":"official","path":"grit/finetuning.py","file_url":"https://github.com/liamma/grit/blob/HEAD/grit/finetuning.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"1bb331bf70a78136"}},{"code_sha256_prefix":"b5c9e7acf599a98a","entry":"get_log_deg","repo":"liamma/grit","repo_kind":"official","path":"grit/layer/grit_layer.py","file_url":"https://github.com/liamma/grit/blob/HEAD/grit/layer/grit_layer.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b5c9e7acf599a98a"}},{"code_sha256_prefix":"9e4db46ba20af98e","entry":"init_model_from_pretrained","repo":"liamma/grit","repo_kind":"official","path":"grit/finetuning.py","file_url":"https://github.com/liamma/grit/blob/HEAD/grit/finetuning.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9e4db46ba20af98e"}},{"code_sha256_prefix":"47e492109e6d308f","entry":"subtoken_cross_entropy","repo":"liamma/grit","repo_kind":"official","path":"grit/loss/subtoken_prediction_loss.py","file_url":"https://github.com/liamma/grit/blob/HEAD/grit/loss/subtoken_prediction_loss.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"47e492109e6d308f"}},{"code_sha256_prefix":"932e8da6cc45ebfc","entry":"FeatureEncoder","repo":"linusbao/MoSE","repo_kind":"listed","path":"GRIT/grit/network/grit_model.py","file_url":"https://github.com/linusbao/MoSE/blob/HEAD/GRIT/grit/network/grit_model.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"932e8da6cc45ebfc"}},{"code_sha256_prefix":"8afca3460aa00ce7","entry":"GritTransformer","repo":"linusbao/MoSE","repo_kind":"listed","path":"GRIT/grit/network/grit_model.py","file_url":"https://github.com/linusbao/MoSE/blob/HEAD/GRIT/grit/network/grit_model.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"8afca3460aa00ce7"}},{"code_sha256_prefix":"dc82c18c9d42c1c1","entry":"adj_l1_losses","repo":"liamma/grit","repo_kind":"official","path":"grit/loss/adj.py","file_url":"https://github.com/liamma/grit/blob/HEAD/grit/loss/adj.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"dc82c18c9d42c1c1"}},{"code_sha256_prefix":"d59c374db2cf64ae","entry":"l1_losses","repo":"liamma/grit","repo_kind":"official","path":"grit/loss/l1.py","file_url":"https://github.com/liamma/grit/blob/HEAD/grit/loss/l1.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d59c374db2cf64ae"}},{"code_sha256_prefix":"91c1f5bb90bbe0f3","entry":"load_pretrained_model_cfg","repo":"liamma/grit","repo_kind":"official","path":"grit/finetuning.py","file_url":"https://github.com/liamma/grit/blob/HEAD/grit/finetuning.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"91c1f5bb90bbe0f3"}},{"code_sha256_prefix":"f1cbbbc857d8b871","entry":"multilabel_cross_entropy","repo":"liamma/grit","repo_kind":"official","path":"grit/loss/multilabel_classification_loss.py","file_url":"https://github.com/liamma/grit/blob/HEAD/grit/loss/multilabel_classification_loss.py","link_basis":"plan_row","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f1cbbbc857d8b871"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}