{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:EFRRHSDOUUUOMGQAIQN7GUAHPN","short_pith_number":"pith:EFRRHSDO","canonical_record":{"source":{"id":"2605.06355","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-07T14:34:08Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"493c5d40d49900c4ca68c54cf476c186904122d586c48e470771c5ce8eb9a929","abstract_canon_sha256":"8ca1ce028131d9ef0234263ebf9c6e483cf3b39a2d5c9c9ab9ec9b984ba9f1d0"},"schema_version":"1.0"},"canonical_sha256":"216313c86ea528e61a00441bf350077b4a2559d78b3820185af1d06ff5d9f57c","source":{"kind":"arxiv","id":"2605.06355","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.06355","created_at":"2026-05-29T02:05:46Z"},{"alias_kind":"arxiv_version","alias_value":"2605.06355v2","created_at":"2026-05-29T02:05:46Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.06355","created_at":"2026-05-29T02:05:46Z"},{"alias_kind":"pith_short_12","alias_value":"EFRRHSDOUUUO","created_at":"2026-05-29T02:05:46Z"},{"alias_kind":"pith_short_16","alias_value":"EFRRHSDOUUUOMGQA","created_at":"2026-05-29T02:05:46Z"},{"alias_kind":"pith_short_8","alias_value":"EFRRHSDO","created_at":"2026-05-29T02:05:46Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:EFRRHSDOUUUOMGQAIQN7GUAHPN","target":"record","payload":{"canonical_record":{"source":{"id":"2605.06355","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-07T14:34:08Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"493c5d40d49900c4ca68c54cf476c186904122d586c48e470771c5ce8eb9a929","abstract_canon_sha256":"8ca1ce028131d9ef0234263ebf9c6e483cf3b39a2d5c9c9ab9ec9b984ba9f1d0"},"schema_version":"1.0"},"canonical_sha256":"216313c86ea528e61a00441bf350077b4a2559d78b3820185af1d06ff5d9f57c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T02:05:46.189519Z","signature_b64":"4zqNOeKHLQZmTt2T+l4QHEtG/aWrImaIiEtObFD4FwLiSnHzMyULNv0ZScVycUWJeOob0VGG78dYakbuoG/vDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"216313c86ea528e61a00441bf350077b4a2559d78b3820185af1d06ff5d9f57c","last_reissued_at":"2026-05-29T02:05:46.188766Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T02:05:46.188766Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.06355","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-29T02:05:46Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"347l9MrqVWC08tHNl5Da23Fq1NxPunTau4PBBT4mMD8gVkulTvNKyRpUGc0MWfBrpR8E0gB2AMg8Gh15mmmqDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-23T14:15:31.727898Z"},"content_sha256":"3aede5f0c6ccd633e0c4a03d6ddd14425828498edd67a46b12c979a1051be746","schema_version":"1.0","event_id":"sha256:3aede5f0c6ccd633e0c4a03d6ddd14425828498edd67a46b12c979a1051be746"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:EFRRHSDOUUUOMGQAIQN7GUAHPN","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Order-Agnostic Autoregressive Modelling with Missing Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Order-agnostic autoregressive models can be trained directly on incomplete data to impute values and choose which ones to query next.","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Ignacio Peis, Jes Frellsen, Pablo M. Olmos","submitted_at":"2026-05-07T14:34:08Z","abstract_excerpt":"Order-Agnostic autoregressive models have demonstrated strong performance in deep generative modeling, yet their use in settings with incomplete data remains largely unexplored. In this work, we reinterpret them through the lens of missing data. First, we show that their standard training procedure on fully observed data implicitly performs imputation under a missing completely at random mechanism, resulting in robust out-of-sample imputation performance in settings with high missingness. Second, we introduce the first principled framework for training them directly on incomplete datasets unde"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Across a suite of real-world benchmarks, our Missingness-Aware Order-Agnostic Autoregressive Model (MO-ARM) consistently outperforms established imputation baselines.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The reinterpretation that standard training on fully observed data implicitly performs imputation under an MCAR mechanism, and that the new framework extends this to general missingness mechanisms without introducing bias, holds as stated in the abstract.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Order-agnostic autoregressive models are extended via a missingness-aware training framework (MO-ARM) that enables direct learning from incomplete data and active information acquisition, outperforming standard imputation methods on real benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Order-agnostic autoregressive models can be trained directly on incomplete data to impute values and choose which ones to query next.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"ff1f7c417f15ebb414d1cb225e75ee48d7ea882e0d1853bfac86d102ebbf7c8b"},"source":{"id":"2605.06355","kind":"arxiv","version":2},"verdict":{"id":"7d2d956f-47ab-4168-9995-d1cdd495d1ad","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T12:54:36.450253Z","strongest_claim":"Across a suite of real-world benchmarks, our Missingness-Aware Order-Agnostic Autoregressive Model (MO-ARM) consistently outperforms established imputation baselines.","one_line_summary":"Order-agnostic autoregressive models are extended via a missingness-aware training framework (MO-ARM) that enables direct learning from incomplete data and active information acquisition, outperforming standard imputation methods on real benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The reinterpretation that standard training on fully observed data implicitly performs imputation under an MCAR mechanism, and that the new framework extends this to general missingness mechanisms without introducing bias, holds as stated in the abstract.","pith_extraction_headline":"Order-agnostic autoregressive models can be trained directly on incomplete data to impute values and choose which ones to query next."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.06355/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-20T12:42:04.021006Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-20T07:43:56.189925Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T18:31:19.322414Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T12:45:04.473758Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"8b0e8de99694b749c6383b6c36d88d626cbdf961f64d751e6d01483587aebfc0"},"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"},"verdict_id":"7d2d956f-47ab-4168-9995-d1cdd495d1ad"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-29T02:05:46Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"vOaHkekxm5X9cH+dqlBWiADyQ9+k9JwnsDAJjs2TxqGfyz2Bkhn8p1L18O0DXkxcDdfAfn+rKEC7uYfZJ1VqCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-23T14:15:31.728417Z"},"content_sha256":"67146474756357a3028ae5ff222afbe0d995a7c48960c21e0c2be319bc87a39b","schema_version":"1.0","event_id":"sha256:67146474756357a3028ae5ff222afbe0d995a7c48960c21e0c2be319bc87a39b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/EFRRHSDOUUUOMGQAIQN7GUAHPN/bundle.json","state_url":"https://pith.science/pith/EFRRHSDOUUUOMGQAIQN7GUAHPN/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/EFRRHSDOUUUOMGQAIQN7GUAHPN/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-23T14:15:31Z","links":{"resolver":"https://pith.science/pith/EFRRHSDOUUUOMGQAIQN7GUAHPN","bundle":"https://pith.science/pith/EFRRHSDOUUUOMGQAIQN7GUAHPN/bundle.json","state":"https://pith.science/pith/EFRRHSDOUUUOMGQAIQN7GUAHPN/state.json","well_known_bundle":"https://pith.science/.well-known/pith/EFRRHSDOUUUOMGQAIQN7GUAHPN/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:EFRRHSDOUUUOMGQAIQN7GUAHPN","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"8ca1ce028131d9ef0234263ebf9c6e483cf3b39a2d5c9c9ab9ec9b984ba9f1d0","cross_cats_sorted":["stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-07T14:34:08Z","title_canon_sha256":"493c5d40d49900c4ca68c54cf476c186904122d586c48e470771c5ce8eb9a929"},"schema_version":"1.0","source":{"id":"2605.06355","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.06355","created_at":"2026-05-29T02:05:46Z"},{"alias_kind":"arxiv_version","alias_value":"2605.06355v2","created_at":"2026-05-29T02:05:46Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.06355","created_at":"2026-05-29T02:05:46Z"},{"alias_kind":"pith_short_12","alias_value":"EFRRHSDOUUUO","created_at":"2026-05-29T02:05:46Z"},{"alias_kind":"pith_short_16","alias_value":"EFRRHSDOUUUOMGQA","created_at":"2026-05-29T02:05:46Z"},{"alias_kind":"pith_short_8","alias_value":"EFRRHSDO","created_at":"2026-05-29T02:05:46Z"}],"graph_snapshots":[{"event_id":"sha256:67146474756357a3028ae5ff222afbe0d995a7c48960c21e0c2be319bc87a39b","target":"graph","created_at":"2026-05-29T02:05:46Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Across a suite of real-world benchmarks, our Missingness-Aware Order-Agnostic Autoregressive Model (MO-ARM) consistently outperforms established imputation baselines."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The reinterpretation that standard training on fully observed data implicitly performs imputation under an MCAR mechanism, and that the new framework extends this to general missingness mechanisms without introducing bias, holds as stated in the abstract."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Order-agnostic autoregressive models are extended via a missingness-aware training framework (MO-ARM) that enables direct learning from incomplete data and active information acquisition, outperforming standard imputation methods on real benchmarks."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Order-agnostic autoregressive models can be trained directly on incomplete data to impute values and choose which ones to query next."}],"snapshot_sha256":"ff1f7c417f15ebb414d1cb225e75ee48d7ea882e0d1853bfac86d102ebbf7c8b"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"claim_evidence","ran_at":"2026-05-20T12:42:04.021006Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-20T07:43:56.189925Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-19T18:31:19.322414Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T12:45:04.473758Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2605.06355/integrity.json","findings":[],"snapshot_sha256":"8b0e8de99694b749c6383b6c36d88d626cbdf961f64d751e6d01483587aebfc0","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Order-Agnostic autoregressive models have demonstrated strong performance in deep generative modeling, yet their use in settings with incomplete data remains largely unexplored. In this work, we reinterpret them through the lens of missing data. First, we show that their standard training procedure on fully observed data implicitly performs imputation under a missing completely at random mechanism, resulting in robust out-of-sample imputation performance in settings with high missingness. Second, we introduce the first principled framework for training them directly on incomplete datasets unde","authors_text":"Ignacio Peis, Jes Frellsen, Pablo M. Olmos","cross_cats":["stat.ML"],"headline":"Order-agnostic autoregressive models can be trained directly on incomplete data to impute values and choose which ones to query next.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-07T14:34:08Z","title":"Order-Agnostic Autoregressive Modelling with Missing Data"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.06355","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-08T12:54:36.450253Z","id":"7d2d956f-47ab-4168-9995-d1cdd495d1ad","model_set":{"reader":"grok-4.3"},"one_line_summary":"Order-agnostic autoregressive models are extended via a missingness-aware training framework (MO-ARM) that enables direct learning from incomplete data and active information acquisition, outperforming standard imputation methods on real benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Order-agnostic autoregressive models can be trained directly on incomplete data to impute values and choose which ones to query next.","strongest_claim":"Across a suite of real-world benchmarks, our Missingness-Aware Order-Agnostic Autoregressive Model (MO-ARM) consistently outperforms established imputation baselines.","weakest_assumption":"The reinterpretation that standard training on fully observed data implicitly performs imputation under an MCAR mechanism, and that the new framework extends this to general missingness mechanisms without introducing bias, holds as stated in the abstract."}},"verdict_id":"7d2d956f-47ab-4168-9995-d1cdd495d1ad"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:3aede5f0c6ccd633e0c4a03d6ddd14425828498edd67a46b12c979a1051be746","target":"record","created_at":"2026-05-29T02:05:46Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"8ca1ce028131d9ef0234263ebf9c6e483cf3b39a2d5c9c9ab9ec9b984ba9f1d0","cross_cats_sorted":["stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-07T14:34:08Z","title_canon_sha256":"493c5d40d49900c4ca68c54cf476c186904122d586c48e470771c5ce8eb9a929"},"schema_version":"1.0","source":{"id":"2605.06355","kind":"arxiv","version":2}},"canonical_sha256":"216313c86ea528e61a00441bf350077b4a2559d78b3820185af1d06ff5d9f57c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"216313c86ea528e61a00441bf350077b4a2559d78b3820185af1d06ff5d9f57c","first_computed_at":"2026-05-29T02:05:46.188766Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-29T02:05:46.188766Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4zqNOeKHLQZmTt2T+l4QHEtG/aWrImaIiEtObFD4FwLiSnHzMyULNv0ZScVycUWJeOob0VGG78dYakbuoG/vDA==","signature_status":"signed_v1","signed_at":"2026-05-29T02:05:46.189519Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.06355","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3aede5f0c6ccd633e0c4a03d6ddd14425828498edd67a46b12c979a1051be746","sha256:67146474756357a3028ae5ff222afbe0d995a7c48960c21e0c2be319bc87a39b"],"state_sha256":"2e477b7485aaaf2adbd8ac13bd123b5a2b80fc26d220b6ace51534d6f26895e2"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MsZuibVQzcb+nbrBhJomvwhCdl4ZpSOi03zizF/79Ms4fxX+xL1YMNu//cmS/2V1lueHuZH/ySu/JaiHDMOuBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-23T14:15:31.732306Z","bundle_sha256":"2808d8a191f01863c7db1ea02cda74fef611160aeea288aecafcd129394a61f4"}}