{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:XSJCDPQS372E5DUFEUGV5RKXKQ","short_pith_number":"pith:XSJCDPQS","canonical_record":{"source":{"id":"2605.13986","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T18:01:43Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"daf73d720c130d23fcec478f4224a2da1974d195d2e321265396253c90e5ec7d","abstract_canon_sha256":"934474672251d10a921e9abdad8961e6b2e4069302316d6df9a0e979a4754a03"},"schema_version":"1.0"},"canonical_sha256":"bc9221be12dff44e8e85250d5ec5575405b13f3d27971128803d2e7670bb5e8a","source":{"kind":"arxiv","id":"2605.13986","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13986","created_at":"2026-05-17T23:39:13Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13986v1","created_at":"2026-05-17T23:39:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13986","created_at":"2026-05-17T23:39:13Z"},{"alias_kind":"pith_short_12","alias_value":"XSJCDPQS372E","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"XSJCDPQS372E5DUF","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"XSJCDPQS","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:XSJCDPQS372E5DUFEUGV5RKXKQ","target":"record","payload":{"canonical_record":{"source":{"id":"2605.13986","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T18:01:43Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"daf73d720c130d23fcec478f4224a2da1974d195d2e321265396253c90e5ec7d","abstract_canon_sha256":"934474672251d10a921e9abdad8961e6b2e4069302316d6df9a0e979a4754a03"},"schema_version":"1.0"},"canonical_sha256":"bc9221be12dff44e8e85250d5ec5575405b13f3d27971128803d2e7670bb5e8a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:13.319843Z","signature_b64":"1F//dGbgfqtVkZoTptkVvYg9C+z8IpQMECyakMLcOvyrrmZQ4rHMT/GQ3rviyGapvJnFp7lyS+gX75/q7zrhBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bc9221be12dff44e8e85250d5ec5575405b13f3d27971128803d2e7670bb5e8a","last_reissued_at":"2026-05-17T23:39:13.319210Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:13.319210Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.13986","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FIlYIcgjRNr8kTdUy5hYiW6T0rJQFe8rKZQIy2swAePBjuylERde5WQWpb4qUlgNzOP5Gl9l8jbe4e5TSjooAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T08:15:36.324263Z"},"content_sha256":"171ba927f6acec67d65988e7d09625ea10460c125e9c51276b76f71a649250bf","schema_version":"1.0","event_id":"sha256:171ba927f6acec67d65988e7d09625ea10460c125e9c51276b76f71a649250bf"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:XSJCDPQS372E5DUFEUGV5RKXKQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"TabPFN-3: Technical Report","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"TabPFN-3 outperforms all tuned and ensembled models on the TabArena tabular benchmark with a single forward pass.","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Adrian Hayler, Alan Arazi, Anurag Garg, Benjamin J\\\"ager, Bernhard Sch\\\"olkopf, Brendan Roof, Clara Cornu, David Salinas, Diana Kriuchkova, Dominik Safaric, Eliott Kalfon, Felix Birkel, Frank Hutter, Georg Grab, Jake Robertson, Jan Hendrik Metzen, Jerry Chen, Julien Siems, Klemens Fl\\\"oge, Kursat Kaya, Lennart Purucker, L\\'eo Grinsztajn, Lilly Charlotte Wehrhahn, Lydia Sidhoum, Madelon Hulsebos, Magnus B\\\"uhler, Marie Salmon, Mihir Manium, Nick Erickson, Noah Hollmann, Oscar Key, Philipp Jund, Philipp Singer, Samuel M\\\"uller, Sauraj Gambhir, Shi Bin (Liam) Hoo, Simon Bing, Simone Alessi, Siyuan Guo, Vladyslav Moroshan, Yann LeCun","submitted_at":"2026-05-13T18:01:43Z","abstract_excerpt":"Tabular data underpins most high-value prediction problems in science and industry, and TabPFN has driven the foundation model revolution for this modality. Designed with feedback from our users, TabPFN-3 builds on this foundation to scale state-of-the-art performance to datasets with 1M training rows and substantially reduce training and inference time. Pretrained exclusively on synthetic data from our prior, TabPFN-3 dramatically pushes the frontier of tabular prediction and brings substantial gains on time series, relational, and tabular-text data. On the standard tabular benchmark TabArena"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"On the standard tabular benchmark TabArena, a forward pass of TabPFN-3 outperforms all other models, including tuned and ensembled baselines, by a significant margin, and pareto-dominates the speed/performance frontier. TabPFN-3-Plus (Thinking) beats all non-TabPFN models by over 200 Elo on TabArena, rising to 420 Elo on the largest data subset, and outperforms AutoGluon 1.5 extreme while being 10x faster.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That benchmark wins on TabArena and other reported datasets, achieved via synthetic pretraining and test-time scaling, will generalize to arbitrary unseen real-world tabular distributions without hidden overfitting to the evaluation suites or synthetic data generator.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"TabPFN-3 delivers state-of-the-art tabular prediction performance on benchmarks up to 1M rows, is up to 20x faster than prior versions, and introduces test-time scaling that beats non-TabPFN models by hundreds of Elo points.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"TabPFN-3 outperforms all tuned and ensembled models on the TabArena tabular benchmark with a single forward pass.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"9c2eb3fb0e63e7af2fed6b1e71e2d46fa1e1dd374ca90cf468c24ae1b34b8618"},"source":{"id":"2605.13986","kind":"arxiv","version":1},"verdict":{"id":"247b931c-a5c8-4309-87e2-c24bc67e4e35","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T06:01:00.864685Z","strongest_claim":"On the standard tabular benchmark TabArena, a forward pass of TabPFN-3 outperforms all other models, including tuned and ensembled baselines, by a significant margin, and pareto-dominates the speed/performance frontier. TabPFN-3-Plus (Thinking) beats all non-TabPFN models by over 200 Elo on TabArena, rising to 420 Elo on the largest data subset, and outperforms AutoGluon 1.5 extreme while being 10x faster.","one_line_summary":"TabPFN-3 delivers state-of-the-art tabular prediction performance on benchmarks up to 1M rows, is up to 20x faster than prior versions, and introduces test-time scaling that beats non-TabPFN models by hundreds of Elo points.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That benchmark wins on TabArena and other reported datasets, achieved via synthetic pretraining and test-time scaling, will generalize to arbitrary unseen real-world tabular distributions without hidden overfitting to the evaluation suites or synthetic data generator.","pith_extraction_headline":"TabPFN-3 outperforms all tuned and ensembled models on the TabArena tabular benchmark with a single forward pass."},"references":{"count":298,"sample":[{"doi":"","year":2025,"title":"arXiv:2506.16791 [cs]","work_id":"155b5349-dee8-4870-965c-d54a700a19de","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2003,"title":"AutoGluon-Tabular: Robust and Accurate AutoML for Structured Data","work_id":"32ca4e6c-bd72-4586-8594-40eb6bcb6582","ref_index":2,"cited_arxiv_id":"2003.06505","is_internal_anchor":true},{"doi":"","year":2015,"title":"A targeted real-time early warning score (trewscore) for septic shock.Science translational medicine, 7(299):299ra122–299ra122, 2015","work_id":"d0ee9103-1d6c-4559-a94d-4a9b0982590b","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2016,"title":"Mimic-iii, a freely accessible critical care database.Scientific data, 3(1):1–9","work_id":"13d40ae1-31af-44be-b076-0f695baf06c1","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"Deep neural networks detect suicide risk from textual facebook posts.Scientific reports, 10(1):16685, 2020","work_id":"c37e433c-b505-4c25-b86d-28a7053c35b7","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":298,"snapshot_sha256":"64be74b4799e08a11111a4ae45246b9dfc0fa4b9b032d5ed15177f39b6103526","internal_anchors":10},"formal_canon":{"evidence_count":2,"snapshot_sha256":"a704e78e85789a12bd7a5c9703756f095b1a5c40476d4b30cf6d9deb1483977e"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"247b931c-a5c8-4309-87e2-c24bc67e4e35"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:39:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RQ29kGGannIGR3tUl4DV2Lyq2nRIbrh/t51tcLvCc/iINDwdnFNl+Xziok8QIUuuot/5JjnsbzvUvgkJfBlMBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T08:15:36.325377Z"},"content_sha256":"d4f243df8dbab91b70dc17a0af32808f85327516a587151b9a0ef5be6363ea0d","schema_version":"1.0","event_id":"sha256:d4f243df8dbab91b70dc17a0af32808f85327516a587151b9a0ef5be6363ea0d"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:XSJCDPQS372E5DUFEUGV5RKXKQ","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.48550/arXiv.2510.06162.URLhttps://arxiv.org/abs/2510.06162) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"Christopher Kolberg, Katharina Eggensperger, and Nico Pfeifer. Tabpfn-wide: Continued pre- training for extreme feature counts.arXiv preprint arXiv:2510.06162, 2025. doi: 10.48550/arXiv. 2510.06162. URLhttps://arxiv.org/abs/2510.06162","arxiv_id":"2605.13986","detector":"doi_compliance","evidence":{"ref_index":181,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"Christopher Kolberg, Katharina Eggensperger, and Nico Pfeifer. Tabpfn-wide: Continued pre- training for extreme feature counts.arXiv preprint arXiv:2510.06162, 2025. doi: 10.48550/arXiv. 2510.06162. URLhttps://arxiv.org/abs/2510.06162","reconstructed_doi":"10.48550/arXiv.2510.06162.URLhttps://arxiv.org/abs/2510.06162"},"severity":"advisory","ref_index":181,"audited_at":"2026-05-19T06:02:08.641333Z","event_type":"pith.integrity.v1","detected_doi":"10.48550/arXiv.2510.06162.URLhttps://arxiv.org/abs/2510.06162","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"ab9a4775eb1f5c8c3ce4784732c8f8596ea818d2276eb020b1187a9ea7df490c","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.0.0","detected_arxiv_id":null,"integrity_event_id":90,"payload_sha256":"1fbca25af0f341764b90c6e1186d28deb1ea0411151a306afc4ebcadb253a56c","signature_b64":"2/7oKXl40Fpy5WiyLGAAHZpgwr9eUTu/9i7G6O2P98TP1fa517kgCWrv6cjk5wrIHRZYyO609zgQy/cV9R9SDA==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-19T06:06:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uZuLYcyO1R+ruAcUGGI9pV1JFLL3oKwvtEzhErRftbaaNSavmHkUCnEFdottnMq16YuPdrzFutAMoPoHZtVsAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T08:15:36.326767Z"},"content_sha256":"136c8e770f3ba7f8ed29f7cc681c3f997580fbda039684ee932d42a526b9128c","schema_version":"1.0","event_id":"sha256:136c8e770f3ba7f8ed29f7cc681c3f997580fbda039684ee932d42a526b9128c"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:XSJCDPQS372E5DUFEUGV5RKXKQ","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.6039/j.issn.1001-0408.2025.19.16.URL) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"Xiaohui Lin, Yujia Wang, Lingling Zhang, and et al. Construction of machine learning clas- sification prediction model for vancomycin blood concentrations based on mimic-iv database. China Pharmacy (ZHONGGUO YAOFANG), 36(19):2448–2453, 2025","arxiv_id":"2605.13986","detector":"doi_compliance","evidence":{"ref_index":179,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"Xiaohui Lin, Yujia Wang, Lingling Zhang, and et al. Construction of machine learning clas- sification prediction model for vancomycin blood concentrations based on mimic-iv database. China Pharmacy (ZHONGGUO YAOFANG), 36(19):2448–2453, 2025","reconstructed_doi":"10.6039/j.issn.1001-0408.2025.19.16.URL"},"severity":"advisory","ref_index":179,"audited_at":"2026-05-19T06:02:08.641333Z","event_type":"pith.integrity.v1","detected_doi":"10.6039/j.issn.1001-0408.2025.19.16.URL","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"002e3d6e3ea8010b531685579a0cc3ea2cd74440124b3829f686084c2394d68d","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.0.0","detected_arxiv_id":null,"integrity_event_id":89,"payload_sha256":"ea51e094be0e8d2915819a88fb304c42d82364535d21bdff9d9bb1d764073d2b","signature_b64":"xFObg8QX5Zad1xtm6VJaSBXWNZNRapa/3Z6E/uZvab+wzkv7SY3b+QptvigHusPO9y8Qtqt2k9ZV2GqrYB+NDg==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-19T06:06:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bquOXOMuG5oWpvfZsuOSAX4GQIwBy6JvCmY60rq0xAUGMOhk4LwIIBGnjgT3u6hmTzI21A+BWmJLJejyaOviCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T08:15:36.327290Z"},"content_sha256":"e8b78f36ddd1b8748d959a80fdfe74ab627750cbdf437b4c3e351e3703f82163","schema_version":"1.0","event_id":"sha256:e8b78f36ddd1b8748d959a80fdfe74ab627750cbdf437b4c3e351e3703f82163"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:XSJCDPQS372E5DUFEUGV5RKXKQ","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.1101/2025.09.17.676674v1) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"Asif Adil et al. Advanced deep learning enables prediction of allogeneic stem cell mobilization success. bioRxiv preprint, 2025. URLhttps://www.biorxiv.org/content/10.1101/2025.09.17. 676674v1","arxiv_id":"2605.13986","detector":"doi_compliance","evidence":{"ref_index":164,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"Asif Adil et al. Advanced deep learning enables prediction of allogeneic stem cell mobilization success. bioRxiv preprint, 2025. URLhttps://www.biorxiv.org/content/10.1101/2025.09.17. 676674v1","reconstructed_doi":"10.1101/2025.09.17.676674v1"},"severity":"advisory","ref_index":164,"audited_at":"2026-05-19T06:02:08.641333Z","event_type":"pith.integrity.v1","detected_doi":"10.1101/2025.09.17.676674v1","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"3ffec96aaaeaefd1efa8c3bedecc680516e34e6bb78e6788c560b26946141b96","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.0.0","detected_arxiv_id":null,"integrity_event_id":88,"payload_sha256":"4b2d793fdfa1d71ec4c9c1791dbc663a4192a00d2ed5ee402fae13170e616d6c","signature_b64":"3VpVLD59Af61uJk7RpGDHw6TsNOe+4lRD1pV7z0AX6Bt8QEpKUxGlGTGmHH7xhlFNs6xAx6akJNG61ve+BvpCQ==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-19T06:06:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"B7ekcmmWohOUAcX9c2no6HKPSkQFBT0O3UbgyoluUUNG+zYHczNhX5zORmCFc7cTAPbVHWOrDWeiGfkcWRR9Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T08:15:36.327790Z"},"content_sha256":"b260493efddc3610e5664ce5078810f9bcb2127b8d9cfc1f26e4efd79cfa492c","schema_version":"1.0","event_id":"sha256:b260493efddc3610e5664ce5078810f9bcb2127b8d9cfc1f26e4efd79cfa492c"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:XSJCDPQS372E5DUFEUGV5RKXKQ","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.1101/2025.09.03.25334985v1.35) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"Vinh Nguyen Dao et al. Early prediction of gestational diabetes using integrated cell-free dna features and omics-derived genetic scores. medRxiv preprint, 2025. URLhttps://www.medrxiv. org/content/10.1101/2025.09.03.25334985v1. 35","arxiv_id":"2605.13986","detector":"doi_compliance","evidence":{"ref_index":158,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"Vinh Nguyen Dao et al. Early prediction of gestational diabetes using integrated cell-free dna features and omics-derived genetic scores. medRxiv preprint, 2025. URLhttps://www.medrxiv. org/content/10.1101/2025.09.03.25334985v1. 35","reconstructed_doi":"10.1101/2025.09.03.25334985v1.35"},"severity":"advisory","ref_index":158,"audited_at":"2026-05-19T06:02:08.641333Z","event_type":"pith.integrity.v1","detected_doi":"10.1101/2025.09.03.25334985v1.35","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"c7331e95f4a98ae5a15d12ad29b8e2e5f8ac0ba46e895f5ee51e88007b683fa3","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.0.0","detected_arxiv_id":null,"integrity_event_id":87,"payload_sha256":"6ca21bb3cd1b2cf56e348fcfd452bacdbd3fa21a50c0dd4052ca5a2af09cd864","signature_b64":"Ye4ojMVa9yC1EfzkZmDZHf5Z6MBJEsQoejJp5hEcHLW/85bdTlKFMabI6PsUrexMHFyREtpw8ja7P7HYqPkhBA==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-19T06:06:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8MJlbssuRzikg+75Y4Z2HBm8kjusEutXFX72PkQMT65MCF1ykvep+wcm7xrzVw5242tACDz0deYAN9o6TTWqAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T08:15:36.328260Z"},"content_sha256":"8b21500da1ae350374afb63b103c4b8542f8c1ee5dfec4a8316bfe1208b3ee58","schema_version":"1.0","event_id":"sha256:8b21500da1ae350374afb63b103c4b8542f8c1ee5dfec4a8316bfe1208b3ee58"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:XSJCDPQS372E5DUFEUGV5RKXKQ","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.48550/arXiv.2505.20685.URLhttps://arxiv.org/abs/2505.20685) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"RosenTing-YingYu, CyrilPicard, andFaezAhmed. Git-bo: High-dimensionalbayesianoptimization with tabular foundation models.arXiv preprint arXiv:2505.20685, 2025. doi: 10.48550/arXiv.2505. 20685. URLhttps://arxiv.org/abs/2505.20685","arxiv_id":"2605.13986","detector":"doi_compliance","evidence":{"ref_index":24,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"RosenTing-YingYu, CyrilPicard, andFaezAhmed. Git-bo: High-dimensionalbayesianoptimization with tabular foundation models.arXiv preprint arXiv:2505.20685, 2025. doi: 10.48550/arXiv.2505. 20685. URLhttps://arxiv.org/abs/2505.20685","reconstructed_doi":"10.48550/arXiv.2505.20685.URLhttps://arxiv.org/abs/2505.20685"},"severity":"advisory","ref_index":24,"audited_at":"2026-05-19T06:02:08.641333Z","event_type":"pith.integrity.v1","detected_doi":"10.48550/arXiv.2505.20685.URLhttps://arxiv.org/abs/2505.20685","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"60a03739b97f22887253375514219953ead52ab1e5f46ae6d4fa9ab6c2e0a534","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.0.0","detected_arxiv_id":null,"integrity_event_id":86,"payload_sha256":"011ffe2545eda16177722b1fb42228530039317ff0d71d2d2eadfb4b1f6459ca","signature_b64":"RvboLzC61GYNH2rkqbkSRSfEVf8Qb6NLazSh8w198EbXXyFyleoSTlB6Mcjw3LAnT2/hbxSp8Em5GU7SdP1WCw==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-19T06:06:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7WRI4mW93XVKdEnbjgm9JomORZh51Jqklxn3BURbzw5GVFBetKYDUGS2/eJJuvgv279HLh/1A6SgSQVKuWCXCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T08:15:36.328813Z"},"content_sha256":"b5681eb096d2a2bbe76fe33a6f63f28c7a32e6dedddbe79ec342425b6e277bc6","schema_version":"1.0","event_id":"sha256:b5681eb096d2a2bbe76fe33a6f63f28c7a32e6dedddbe79ec342425b6e277bc6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XSJCDPQS372E5DUFEUGV5RKXKQ/bundle.json","state_url":"https://pith.science/pith/XSJCDPQS372E5DUFEUGV5RKXKQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XSJCDPQS372E5DUFEUGV5RKXKQ/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":"2026-05-21T08:15:36Z","links":{"resolver":"https://pith.science/pith/XSJCDPQS372E5DUFEUGV5RKXKQ","bundle":"https://pith.science/pith/XSJCDPQS372E5DUFEUGV5RKXKQ/bundle.json","state":"https://pith.science/pith/XSJCDPQS372E5DUFEUGV5RKXKQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XSJCDPQS372E5DUFEUGV5RKXKQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:XSJCDPQS372E5DUFEUGV5RKXKQ","merge_version":"pith-open-graph-merge-v1","event_count":7,"valid_event_count":7,"invalid_event_count":0,"equivocation_count":1,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"934474672251d10a921e9abdad8961e6b2e4069302316d6df9a0e979a4754a03","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T18:01:43Z","title_canon_sha256":"daf73d720c130d23fcec478f4224a2da1974d195d2e321265396253c90e5ec7d"},"schema_version":"1.0","source":{"id":"2605.13986","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13986","created_at":"2026-05-17T23:39:13Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13986v1","created_at":"2026-05-17T23:39:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13986","created_at":"2026-05-17T23:39:13Z"},{"alias_kind":"pith_short_12","alias_value":"XSJCDPQS372E","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"XSJCDPQS372E5DUF","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"XSJCDPQS","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:d4f243df8dbab91b70dc17a0af32808f85327516a587151b9a0ef5be6363ea0d","target":"graph","created_at":"2026-05-17T23:39:13Z","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":"On the standard tabular benchmark TabArena, a forward pass of TabPFN-3 outperforms all other models, including tuned and ensembled baselines, by a significant margin, and pareto-dominates the speed/performance frontier. TabPFN-3-Plus (Thinking) beats all non-TabPFN models by over 200 Elo on TabArena, rising to 420 Elo on the largest data subset, and outperforms AutoGluon 1.5 extreme while being 10x faster."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That benchmark wins on TabArena and other reported datasets, achieved via synthetic pretraining and test-time scaling, will generalize to arbitrary unseen real-world tabular distributions without hidden overfitting to the evaluation suites or synthetic data generator."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"TabPFN-3 delivers state-of-the-art tabular prediction performance on benchmarks up to 1M rows, is up to 20x faster than prior versions, and introduces test-time scaling that beats non-TabPFN models by hundreds of Elo points."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"TabPFN-3 outperforms all tuned and ensembled models on the TabArena tabular benchmark with a single forward pass."}],"snapshot_sha256":"9c2eb3fb0e63e7af2fed6b1e71e2d46fa1e1dd374ca90cf468c24ae1b34b8618"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"a704e78e85789a12bd7a5c9703756f095b1a5c40476d4b30cf6d9deb1483977e"},"paper":{"abstract_excerpt":"Tabular data underpins most high-value prediction problems in science and industry, and TabPFN has driven the foundation model revolution for this modality. Designed with feedback from our users, TabPFN-3 builds on this foundation to scale state-of-the-art performance to datasets with 1M training rows and substantially reduce training and inference time. Pretrained exclusively on synthetic data from our prior, TabPFN-3 dramatically pushes the frontier of tabular prediction and brings substantial gains on time series, relational, and tabular-text data. On the standard tabular benchmark TabArena","authors_text":"Adrian Hayler, Alan Arazi, Anurag Garg, Benjamin J\\\"ager, Bernhard Sch\\\"olkopf, Brendan Roof, Clara Cornu, David Salinas, Diana Kriuchkova, Dominik Safaric, Eliott Kalfon, Felix Birkel, Frank Hutter, Georg Grab, Jake Robertson, Jan Hendrik Metzen, Jerry Chen, Julien Siems, Klemens Fl\\\"oge, Kursat Kaya, Lennart Purucker, L\\'eo Grinsztajn, Lilly Charlotte Wehrhahn, Lydia Sidhoum, Madelon Hulsebos, Magnus B\\\"uhler, Marie Salmon, Mihir Manium, Nick Erickson, Noah Hollmann, Oscar Key, Philipp Jund, Philipp Singer, Samuel M\\\"uller, Sauraj Gambhir, Shi Bin (Liam) Hoo, Simon Bing, Simone Alessi, Siyuan Guo, Vladyslav Moroshan, Yann LeCun","cross_cats":["stat.ML"],"headline":"TabPFN-3 outperforms all tuned and ensembled models on the TabArena tabular benchmark with a single forward pass.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T18:01:43Z","title":"TabPFN-3: Technical Report"},"references":{"count":298,"internal_anchors":10,"resolved_work":298,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"arXiv:2506.16791 [cs]","work_id":"155b5349-dee8-4870-965c-d54a700a19de","year":2025},{"cited_arxiv_id":"2003.06505","doi":"","is_internal_anchor":true,"ref_index":2,"title":"AutoGluon-Tabular: Robust and Accurate AutoML for Structured Data","work_id":"32ca4e6c-bd72-4586-8594-40eb6bcb6582","year":2003},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"A targeted real-time early warning score (trewscore) for septic shock.Science translational medicine, 7(299):299ra122–299ra122, 2015","work_id":"d0ee9103-1d6c-4559-a94d-4a9b0982590b","year":2015},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Mimic-iii, a freely accessible critical care database.Scientific data, 3(1):1–9","work_id":"13d40ae1-31af-44be-b076-0f695baf06c1","year":2016},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Deep neural networks detect suicide risk from textual facebook posts.Scientific reports, 10(1):16685, 2020","work_id":"c37e433c-b505-4c25-b86d-28a7053c35b7","year":2020}],"snapshot_sha256":"64be74b4799e08a11111a4ae45246b9dfc0fa4b9b032d5ed15177f39b6103526"},"source":{"id":"2605.13986","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-15T06:01:00.864685Z","id":"247b931c-a5c8-4309-87e2-c24bc67e4e35","model_set":{"reader":"grok-4.3"},"one_line_summary":"TabPFN-3 delivers state-of-the-art tabular prediction performance on benchmarks up to 1M rows, is up to 20x faster than prior versions, and introduces test-time scaling that beats non-TabPFN models by hundreds of Elo points.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"TabPFN-3 outperforms all tuned and ensembled models on the TabArena tabular benchmark with a single forward pass.","strongest_claim":"On the standard tabular benchmark TabArena, a forward pass of TabPFN-3 outperforms all other models, including tuned and ensembled baselines, by a significant margin, and pareto-dominates the speed/performance frontier. TabPFN-3-Plus (Thinking) beats all non-TabPFN models by over 200 Elo on TabArena, rising to 420 Elo on the largest data subset, and outperforms AutoGluon 1.5 extreme while being 10x faster.","weakest_assumption":"That benchmark wins on TabArena and other reported datasets, achieved via synthetic pretraining and test-time scaling, will generalize to arbitrary unseen real-world tabular distributions without hidden overfitting to the evaluation suites or synthetic data generator."}},"verdict_id":"247b931c-a5c8-4309-87e2-c24bc67e4e35"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:171ba927f6acec67d65988e7d09625ea10460c125e9c51276b76f71a649250bf","target":"record","created_at":"2026-05-17T23:39:13Z","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":"934474672251d10a921e9abdad8961e6b2e4069302316d6df9a0e979a4754a03","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T18:01:43Z","title_canon_sha256":"daf73d720c130d23fcec478f4224a2da1974d195d2e321265396253c90e5ec7d"},"schema_version":"1.0","source":{"id":"2605.13986","kind":"arxiv","version":1}},"canonical_sha256":"bc9221be12dff44e8e85250d5ec5575405b13f3d27971128803d2e7670bb5e8a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bc9221be12dff44e8e85250d5ec5575405b13f3d27971128803d2e7670bb5e8a","first_computed_at":"2026-05-17T23:39:13.319210Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:39:13.319210Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"1F//dGbgfqtVkZoTptkVvYg9C+z8IpQMECyakMLcOvyrrmZQ4rHMT/GQ3rviyGapvJnFp7lyS+gX75/q7zrhBw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:39:13.319843Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13986","source_kind":"arxiv","source_version":1}}},"equivocations":[{"signer_id":"pith.science","event_type":"integrity_finding","target":"integrity","event_ids":["sha256:136c8e770f3ba7f8ed29f7cc681c3f997580fbda039684ee932d42a526b9128c","sha256:8b21500da1ae350374afb63b103c4b8542f8c1ee5dfec4a8316bfe1208b3ee58","sha256:b260493efddc3610e5664ce5078810f9bcb2127b8d9cfc1f26e4efd79cfa492c","sha256:b5681eb096d2a2bbe76fe33a6f63f28c7a32e6dedddbe79ec342425b6e277bc6","sha256:e8b78f36ddd1b8748d959a80fdfe74ab627750cbdf437b4c3e351e3703f82163"]}],"invalid_events":[],"applied_event_ids":["sha256:171ba927f6acec67d65988e7d09625ea10460c125e9c51276b76f71a649250bf","sha256:d4f243df8dbab91b70dc17a0af32808f85327516a587151b9a0ef5be6363ea0d"],"state_sha256":"3a8caa3fdb321628715c1e4be2c54aaf60de1d70637744cf46d0c1329489b316"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/UMnP0sQKtlZ7TlZA+evP7y4MH22aMovbW3wQIfp9PK0+LqQX0RBRnXkxjEA0WK4ZNzJ95ezMI+88F/6s/pQDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-21T08:15:36.334444Z","bundle_sha256":"ffa5d205e6278467abbbee2d1f3b1334cd56e4a2ba62374c5c600e56e706e164"}}