{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:R2NGLJMHATMRFN6CQLRGUTVVLW","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":"8915f57284c5cbe413b46be146a082223f26cbda7b4c8f54319362bb47934390","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2019-09-04T17:33:42Z","title_canon_sha256":"4f230ff925374b3e919226189dfa72ce53e871c9d1c4f98a20e906de4f59056d"},"schema_version":"1.0","source":{"id":"1909.01958","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1909.01958","created_at":"2026-07-05T02:11:19Z"},{"alias_kind":"arxiv_version","alias_value":"1909.01958v3","created_at":"2026-07-05T02:11:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1909.01958","created_at":"2026-07-05T02:11:19Z"},{"alias_kind":"pith_short_12","alias_value":"R2NGLJMHATMR","created_at":"2026-07-05T02:11:19Z"},{"alias_kind":"pith_short_16","alias_value":"R2NGLJMHATMRFN6C","created_at":"2026-07-05T02:11:19Z"},{"alias_kind":"pith_short_8","alias_value":"R2NGLJMH","created_at":"2026-07-05T02:11:19Z"}],"graph_snapshots":[{"event_id":"sha256:0c6b5ba43db31337c68c08a3d772d4a0b984fd3dfce0843026eb7442d09ece66","target":"graph","created_at":"2026-07-05T02:11:19Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/1909.01958/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"AI has achieved remarkable mastery over games such as Chess, Go, and Poker, and even Jeopardy, but the rich variety of standardized exams has remained a landmark challenge. Even in 2016, the best AI system achieved merely 59.3% on an 8th Grade science exam challenge. This paper reports unprecedented success on the Grade 8 New York Regents Science Exam, where for the first time a system scores more than 90% on the exam's non-diagram, multiple choice (NDMC) questions. In addition, our Aristo system, building upon the success of recent language models, exceeded 83% on the corresponding Grade 12 S","authors_text":"Ashish Sabharwal, Bhavana Dalvi Mishra, Carissa Schoenick, Daniel Khashabi, Dirk Groeneveld, Kyle Richardson, Michael Schmitz, Michal Guerquin, Niket Tandon, Oren Etzioni, Oyvind Tafjord, Peter Clark, Sumithra Bhakthavatsalam, Tushar Khot","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2019-09-04T17:33:42Z","title":"From 'F' to 'A' on the N.Y. Regents Science Exams: An Overview of the Aristo Project"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1909.01958","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:1868f0651ba4f75c00637f8501efb350e6568c85be455f3fa99fc74b33b43af7","target":"record","created_at":"2026-07-05T02:11:19Z","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":"8915f57284c5cbe413b46be146a082223f26cbda7b4c8f54319362bb47934390","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2019-09-04T17:33:42Z","title_canon_sha256":"4f230ff925374b3e919226189dfa72ce53e871c9d1c4f98a20e906de4f59056d"},"schema_version":"1.0","source":{"id":"1909.01958","kind":"arxiv","version":3}},"canonical_sha256":"8e9a65a58704d912b7c282e26a4eb55d89e2f28ce7504213bb7d50d569f9344b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8e9a65a58704d912b7c282e26a4eb55d89e2f28ce7504213bb7d50d569f9344b","first_computed_at":"2026-07-05T02:11:19.820058Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T02:11:19.820058Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"NCqLol3oMrbzDQ5yBWo4+S08K9BkKVMje5BJr3xpH18i/crJxsPZTudKfn5EpLSXCva94UtK/UUj0mZdxSR3Aw==","signature_status":"signed_v1","signed_at":"2026-07-05T02:11:19.820503Z","signed_message":"canonical_sha256_bytes"},"source_id":"1909.01958","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1868f0651ba4f75c00637f8501efb350e6568c85be455f3fa99fc74b33b43af7","sha256:0c6b5ba43db31337c68c08a3d772d4a0b984fd3dfce0843026eb7442d09ece66"],"state_sha256":"18555da187d755ef81e01852523bfbd01943f000162830b2cf4c2a6a652a5e46"}