{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:A67FZDI7OWUQDTEUJKEES5NB75","short_pith_number":"pith:A67FZDI7","schema_version":"1.0","canonical_sha256":"07be5c8d1f75a901cc944a884975a1ff7a5d7e18dea4e2e544a5b5a665116e66","source":{"kind":"arxiv","id":"2506.22084","version":1},"attestation_state":"computed","paper":{"title":"Transformers are Graph Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Chaitanya K. Joshi","submitted_at":"2025-06-27T10:15:33Z","abstract_excerpt":"We establish connections between the Transformer architecture, originally introduced for natural language processing, and Graph Neural Networks (GNNs) for representation learning on graphs. We show how Transformers can be viewed as message passing GNNs operating on fully connected graphs of tokens, where the self-attention mechanism capture the relative importance of all tokens w.r.t. each-other, and positional encodings provide hints about sequential ordering or structure. Thus, Transformers are expressive set processing networks that learn relationships among input elements without being con"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2506.22084","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-06-27T10:15:33Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5171957ca685bb8e6a67448e5c26f8328cab9315d4909937be2b823259b26b51","abstract_canon_sha256":"0186146449548b5973f9506a80f285c0f452a53746522fe0ebdfb0b83f4e389b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:28:21.760043Z","signature_b64":"d1vKO9XNvT+Bov7tfhmaT7txm8bhyo02zNji0wlPOnvYRg/GHH4APsDsZfxQ8M6G5gG3dwwmGtoZLTnUKGh4Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"07be5c8d1f75a901cc944a884975a1ff7a5d7e18dea4e2e544a5b5a665116e66","last_reissued_at":"2026-07-05T11:28:21.759337Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:28:21.759337Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Transformers are Graph Neural Networks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Chaitanya K. Joshi","submitted_at":"2025-06-27T10:15:33Z","abstract_excerpt":"We establish connections between the Transformer architecture, originally introduced for natural language processing, and Graph Neural Networks (GNNs) for representation learning on graphs. We show how Transformers can be viewed as message passing GNNs operating on fully connected graphs of tokens, where the self-attention mechanism capture the relative importance of all tokens w.r.t. each-other, and positional encodings provide hints about sequential ordering or structure. Thus, Transformers are expressive set processing networks that learn relationships among input elements without being con"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.22084","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2506.22084/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2506.22084","created_at":"2026-07-05T11:28:21.759400+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.22084v1","created_at":"2026-07-05T11:28:21.759400+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.22084","created_at":"2026-07-05T11:28:21.759400+00:00"},{"alias_kind":"pith_short_12","alias_value":"A67FZDI7OWUQ","created_at":"2026-07-05T11:28:21.759400+00:00"},{"alias_kind":"pith_short_16","alias_value":"A67FZDI7OWUQDTEU","created_at":"2026-07-05T11:28:21.759400+00:00"},{"alias_kind":"pith_short_8","alias_value":"A67FZDI7","created_at":"2026-07-05T11:28:21.759400+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.04833","citing_title":"Signed Dual Attention: Capturing Signed Dependencies in Time Series Forecasting","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2509.22343","citing_title":"Transformers Can Learn Connectivity in Some Graphs but Not Others","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12512","citing_title":"Beyond Individual Mimicry: Constructing Human-Like Social network with Graph-Augmented LLM Agents","ref_index":76,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/A67FZDI7OWUQDTEUJKEES5NB75","json":"https://pith.science/pith/A67FZDI7OWUQDTEUJKEES5NB75.json","graph_json":"https://pith.science/api/pith-number/A67FZDI7OWUQDTEUJKEES5NB75/graph.json","events_json":"https://pith.science/api/pith-number/A67FZDI7OWUQDTEUJKEES5NB75/events.json","paper":"https://pith.science/paper/A67FZDI7"},"agent_actions":{"view_html":"https://pith.science/pith/A67FZDI7OWUQDTEUJKEES5NB75","download_json":"https://pith.science/pith/A67FZDI7OWUQDTEUJKEES5NB75.json","view_paper":"https://pith.science/paper/A67FZDI7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.22084&json=true","fetch_graph":"https://pith.science/api/pith-number/A67FZDI7OWUQDTEUJKEES5NB75/graph.json","fetch_events":"https://pith.science/api/pith-number/A67FZDI7OWUQDTEUJKEES5NB75/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/A67FZDI7OWUQDTEUJKEES5NB75/action/timestamp_anchor","attest_storage":"https://pith.science/pith/A67FZDI7OWUQDTEUJKEES5NB75/action/storage_attestation","attest_author":"https://pith.science/pith/A67FZDI7OWUQDTEUJKEES5NB75/action/author_attestation","sign_citation":"https://pith.science/pith/A67FZDI7OWUQDTEUJKEES5NB75/action/citation_signature","submit_replication":"https://pith.science/pith/A67FZDI7OWUQDTEUJKEES5NB75/action/replication_record"}},"created_at":"2026-07-05T11:28:21.759400+00:00","updated_at":"2026-07-05T11:28:21.759400+00:00"}