{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:W7ZVCLE22AVKXHL7VKSASRGIPE","short_pith_number":"pith:W7ZVCLE2","schema_version":"1.0","canonical_sha256":"b7f3512c9ad02aab9d7faaa40944c879097a9238fd4677c1fb901190c822b822","source":{"kind":"arxiv","id":"2605.13383","version":1},"attestation_state":"computed","paper":{"title":"Beyond Oversquashing: Understanding Signal Propagation in GNNs Via Observables","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Standard spectral GNNs diffuse signals broadly instead of routing them to specific regions, while Schrödinger GNNs maintain better directed propagation.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Eden Nagar, Ron Levie, Ya-Wei Eileen Lin","submitted_at":"2026-05-13T11:38:35Z","abstract_excerpt":"Graph Neural Networks (GNNs) perform computations on graphs by routing the signal between graph regions using a graph shift operator or a message passing scheme. Often, the propagation of the signal leads to a loss of information, where the signal tends to diffuse across the graph instead of being deliberately routed between regions of interest. Two notions that depict this phenomenon are oversmoothing and oversquashing. In this paper, we propose an alternative approach for modeling signal propagation, inspired by quantum mechanics, using the notion of observables. Specifically, we model the p"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2605.13383","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T11:38:35Z","cross_cats_sorted":[],"title_canon_sha256":"de105de8017823c6187a4faaacdb4d9ff19d527a74759c32f065756f6a3aeeff","abstract_canon_sha256":"b752c8442ff87d15a1d0de3ac7a0b50938aa40c01c7ffb78850565773b37c607"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:44:47.812181Z","signature_b64":"8AlmNsNHPJ9+gci5oNu9ZGWZM7IDTnysY+uMIAEle7GihZj5yS3+34CGMf+r/jvdAaidorldhm9MjfhGIvOQAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b7f3512c9ad02aab9d7faaa40944c879097a9238fd4677c1fb901190c822b822","last_reissued_at":"2026-05-18T02:44:47.811738Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:44:47.811738Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Oversquashing: Understanding Signal Propagation in GNNs Via Observables","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Standard spectral GNNs diffuse signals broadly instead of routing them to specific regions, while Schrödinger GNNs maintain better directed propagation.","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Eden Nagar, Ron Levie, Ya-Wei Eileen Lin","submitted_at":"2026-05-13T11:38:35Z","abstract_excerpt":"Graph Neural Networks (GNNs) perform computations on graphs by routing the signal between graph regions using a graph shift operator or a message passing scheme. Often, the propagation of the signal leads to a loss of information, where the signal tends to diffuse across the graph instead of being deliberately routed between regions of interest. Two notions that depict this phenomenon are oversmoothing and oversquashing. In this paper, we propose an alternative approach for modeling signal propagation, inspired by quantum mechanics, using the notion of observables. Specifically, we model the p"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We prove that standard spectral GNNs have poor signal propagation capabilities. We then propose a new type of spectral GNN, termed Schrödinger GNN, which we show has a superior capacity to route the signal across the graph.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the quantum-mechanics-inspired definitions of signal location, concentration, and propagation to a location of interest faithfully capture the discrete, finite-layer behavior of message passing in GNNs.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Quantum-inspired observables reveal poor signal routing in standard spectral GNNs and motivate Schrödinger GNNs with superior propagation capacity.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Standard spectral GNNs diffuse signals broadly instead of routing them to specific regions, while Schrödinger GNNs maintain better directed propagation.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"c70f1d0c318ee48d7f99017f21a4b497bddce329a7e3f700dea04d34c83d96c5"},"source":{"id":"2605.13383","kind":"arxiv","version":1},"verdict":{"id":"18b11463-027c-485b-84be-d41d2668091e","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T20:28:39.859604Z","strongest_claim":"We prove that standard spectral GNNs have poor signal propagation capabilities. We then propose a new type of spectral GNN, termed Schrödinger GNN, which we show has a superior capacity to route the signal across the graph.","one_line_summary":"Quantum-inspired observables reveal poor signal routing in standard spectral GNNs and motivate Schrödinger GNNs with superior propagation capacity.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the quantum-mechanics-inspired definitions of signal location, concentration, and propagation to a location of interest faithfully capture the discrete, finite-layer behavior of message passing in GNNs.","pith_extraction_headline":"Standard spectral GNNs diffuse signals broadly instead of routing them to specific regions, while Schrödinger GNNs maintain better directed propagation."},"references":{"count":20,"sample":[{"doi":"","year":null,"title":"Oversquashing in gnns through the lens of information contraction and graph expansion","work_id":"ca512d05-903f-4dab-8773-8b155a5498e1","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2012,"title":"A generalization of transformer networks to graphs","work_id":"0438ea98-886e-4e29-ac3c-fe11ffe8b0f3","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1002/mana.202100466","year":null,"title":"doi: 10.1002/mana.202100466","work_id":"222afc5c-f5fd-4613-8e9a-627284828211","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Adam: A Method for Stochastic Optimization","work_id":"1910796d-9b52-4683-bf5c-de9632c1028b","ref_index":4,"cited_arxiv_id":"1412.6980","is_internal_anchor":true},{"doi":"","year":null,"title":"Semi-Supervised Classification with Graph Convolutional Networks","work_id":"21fff118-807d-49cd-8229-f7087ba57b5d","ref_index":5,"cited_arxiv_id":"1609.02907","is_internal_anchor":true}],"resolved_work":20,"snapshot_sha256":"7d0f1d07f041c2d82b5ca9c4a26dde9227daf5b6eb0d12bfe77d87bc252f7279","internal_anchors":2},"formal_canon":{"evidence_count":2,"snapshot_sha256":"330e54f5ede8f94bc1acce1604b3f240e66fa6f4b8817568b87450708943dddc"},"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":"2605.13383","created_at":"2026-05-18T02:44:47.811810+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.13383v1","created_at":"2026-05-18T02:44:47.811810+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13383","created_at":"2026-05-18T02:44:47.811810+00:00"},{"alias_kind":"pith_short_12","alias_value":"W7ZVCLE22AVK","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"W7ZVCLE22AVKXHL7","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"W7ZVCLE2","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/W7ZVCLE22AVKXHL7VKSASRGIPE","json":"https://pith.science/pith/W7ZVCLE22AVKXHL7VKSASRGIPE.json","graph_json":"https://pith.science/api/pith-number/W7ZVCLE22AVKXHL7VKSASRGIPE/graph.json","events_json":"https://pith.science/api/pith-number/W7ZVCLE22AVKXHL7VKSASRGIPE/events.json","paper":"https://pith.science/paper/W7ZVCLE2"},"agent_actions":{"view_html":"https://pith.science/pith/W7ZVCLE22AVKXHL7VKSASRGIPE","download_json":"https://pith.science/pith/W7ZVCLE22AVKXHL7VKSASRGIPE.json","view_paper":"https://pith.science/paper/W7ZVCLE2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.13383&json=true","fetch_graph":"https://pith.science/api/pith-number/W7ZVCLE22AVKXHL7VKSASRGIPE/graph.json","fetch_events":"https://pith.science/api/pith-number/W7ZVCLE22AVKXHL7VKSASRGIPE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/W7ZVCLE22AVKXHL7VKSASRGIPE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/W7ZVCLE22AVKXHL7VKSASRGIPE/action/storage_attestation","attest_author":"https://pith.science/pith/W7ZVCLE22AVKXHL7VKSASRGIPE/action/author_attestation","sign_citation":"https://pith.science/pith/W7ZVCLE22AVKXHL7VKSASRGIPE/action/citation_signature","submit_replication":"https://pith.science/pith/W7ZVCLE22AVKXHL7VKSASRGIPE/action/replication_record"}},"created_at":"2026-05-18T02:44:47.811810+00:00","updated_at":"2026-05-18T02:44:47.811810+00:00"}