{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ECNNCKSIKKDU6PU3MOSY7VJWWJ","short_pith_number":"pith:ECNNCKSI","schema_version":"1.0","canonical_sha256":"209ad12a4852874f3e9b63a58fd536b26f8e5996ccddf63f93c259ff4ebccb03","source":{"kind":"arxiv","id":"2503.05979","version":2},"attestation_state":"computed","paper":{"title":"Learning-Order Autoregressive Models with Application to Molecular Graph Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Arthur Gretton, Jiaxin Shi, Michalis K. Titsias, Nicolas Heess, Zhe Wang","submitted_at":"2025-03-07T23:24:24Z","abstract_excerpt":"Autoregressive models (ARMs) have become the workhorse for sequence generation tasks, since many problems can be modeled as next-token prediction. While there appears to be a natural ordering for text (i.e., left-to-right), for many data types, such as graphs, the canonical ordering is less obvious. To address this problem, we introduce a variant of ARM that generates high-dimensional data using a probabilistic ordering that is sequentially inferred from data. This model incorporates a trainable probability distribution, referred to as an order-policy, that dynamically decides the autoregressi"},"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":"2503.05979","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-03-07T23:24:24Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"9dc2e960724abebe9476079f7b9e603868637352766b73c3c0d3478b6d0c6f2a","abstract_canon_sha256":"2392b22d03b5fdc9dd1f84a8ee961af1d380e50f09462dd99ede63d92c15f383"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:35:54.168093Z","signature_b64":"2b0et/XvlUF9d/aTsxq3IqWvYG2n6LeQINUICltnlzNirO82L/E+LnQ6yUp8qpThFlrUxn5eVLJBdYtD4yF5Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"209ad12a4852874f3e9b63a58fd536b26f8e5996ccddf63f93c259ff4ebccb03","last_reissued_at":"2026-07-05T11:35:54.167649Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:35:54.167649Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning-Order Autoregressive Models with Application to Molecular Graph Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Arthur Gretton, Jiaxin Shi, Michalis K. Titsias, Nicolas Heess, Zhe Wang","submitted_at":"2025-03-07T23:24:24Z","abstract_excerpt":"Autoregressive models (ARMs) have become the workhorse for sequence generation tasks, since many problems can be modeled as next-token prediction. While there appears to be a natural ordering for text (i.e., left-to-right), for many data types, such as graphs, the canonical ordering is less obvious. To address this problem, we introduce a variant of ARM that generates high-dimensional data using a probabilistic ordering that is sequentially inferred from data. This model incorporates a trainable probability distribution, referred to as an order-policy, that dynamically decides the autoregressi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.05979","kind":"arxiv","version":2},"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/2503.05979/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":"2503.05979","created_at":"2026-07-05T11:35:54.167706+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.05979v2","created_at":"2026-07-05T11:35:54.167706+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.05979","created_at":"2026-07-05T11:35:54.167706+00:00"},{"alias_kind":"pith_short_12","alias_value":"ECNNCKSIKKDU","created_at":"2026-07-05T11:35:54.167706+00:00"},{"alias_kind":"pith_short_16","alias_value":"ECNNCKSIKKDU6PU3","created_at":"2026-07-05T11:35:54.167706+00:00"},{"alias_kind":"pith_short_8","alias_value":"ECNNCKSI","created_at":"2026-07-05T11:35:54.167706+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.01775","citing_title":"Set Diffusion: Interpolating Token Orderings Between Autoregression and Diffusion for Fast and Flexible Decoding","ref_index":125,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00295","citing_title":"Adaptive Order Policies for Masked Diffusion","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18204","citing_title":"Forward-Learned Discrete Diffusion: Learning how to noise to denoise faster","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05929","citing_title":"ReLU Networks for Exact Generation of Similar Graphs","ref_index":42,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ECNNCKSIKKDU6PU3MOSY7VJWWJ","json":"https://pith.science/pith/ECNNCKSIKKDU6PU3MOSY7VJWWJ.json","graph_json":"https://pith.science/api/pith-number/ECNNCKSIKKDU6PU3MOSY7VJWWJ/graph.json","events_json":"https://pith.science/api/pith-number/ECNNCKSIKKDU6PU3MOSY7VJWWJ/events.json","paper":"https://pith.science/paper/ECNNCKSI"},"agent_actions":{"view_html":"https://pith.science/pith/ECNNCKSIKKDU6PU3MOSY7VJWWJ","download_json":"https://pith.science/pith/ECNNCKSIKKDU6PU3MOSY7VJWWJ.json","view_paper":"https://pith.science/paper/ECNNCKSI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.05979&json=true","fetch_graph":"https://pith.science/api/pith-number/ECNNCKSIKKDU6PU3MOSY7VJWWJ/graph.json","fetch_events":"https://pith.science/api/pith-number/ECNNCKSIKKDU6PU3MOSY7VJWWJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ECNNCKSIKKDU6PU3MOSY7VJWWJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ECNNCKSIKKDU6PU3MOSY7VJWWJ/action/storage_attestation","attest_author":"https://pith.science/pith/ECNNCKSIKKDU6PU3MOSY7VJWWJ/action/author_attestation","sign_citation":"https://pith.science/pith/ECNNCKSIKKDU6PU3MOSY7VJWWJ/action/citation_signature","submit_replication":"https://pith.science/pith/ECNNCKSIKKDU6PU3MOSY7VJWWJ/action/replication_record"}},"created_at":"2026-07-05T11:35:54.167706+00:00","updated_at":"2026-07-05T11:35:54.167706+00:00"}