{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:5N2N3GXHAVGJ3DIAERBA2ED736","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":"927c4889782be6b6bd5c205399e62f8e4869cac9638c87bfc36eb6830589af43","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-05-19T23:57:23Z","title_canon_sha256":"e9e53d4b2c8e24587b3aa33ee5bfedbfafdf21af4617107e422e70fe02959910"},"schema_version":"1.0","source":{"id":"1905.07830","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1905.07830","created_at":"2026-07-04T23:38:28Z"},{"alias_kind":"arxiv_version","alias_value":"1905.07830v1","created_at":"2026-07-04T23:38:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1905.07830","created_at":"2026-07-04T23:38:28Z"},{"alias_kind":"pith_short_12","alias_value":"5N2N3GXHAVGJ","created_at":"2026-07-04T23:38:28Z"},{"alias_kind":"pith_short_16","alias_value":"5N2N3GXHAVGJ3DIA","created_at":"2026-07-04T23:38:28Z"},{"alias_kind":"pith_short_8","alias_value":"5N2N3GXH","created_at":"2026-07-04T23:38:28Z"}],"graph_snapshots":[{"event_id":"sha256:625a6428af8057ed1f663da335021806c87febdda38467521288fc7092efdceb","target":"graph","created_at":"2026-07-04T23:38:28Z","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":"Though its questions are trivial for humans (>95% accuracy), state-of-the-art models struggle (<48%)."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That adversarial filtering produces examples requiring genuine commonsense reasoning rather than merely exploiting specific weaknesses or distributional artifacts in the models used during filtering."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"HellaSwag dataset shows state-of-the-art models fail commonsense inference tasks that humans solve easily, built via adversarial filtering of distractors."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"HellaSwag shows state-of-the-art models still fail at commonsense sentence completion that humans solve easily."}],"snapshot_sha256":"0aba65de350fc7c765ee8f342d88f717b5e1797142ba6879a16035483c3f1f3c"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/1905.07830/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent work by Zellers et al. (2018) introduced a new task of commonsense natural language inference: given an event description such as \"A woman sits at a piano,\" a machine must select the most likely followup: \"She sets her fingers on the keys.\" With the introduction of BERT, near human-level performance was reached. Does this mean that machines can perform human level commonsense inference?\n  In this paper, we show that commonsense inference still proves difficult for even state-of-the-art models, by presenting HellaSwag, a new challenge dataset. Though its questions are trivial for humans ","authors_text":"Ali Farhadi, Ari Holtzman, Rowan Zellers, Yejin Choi, Yonatan Bisk","cross_cats":[],"headline":"HellaSwag shows state-of-the-art models still fail at commonsense sentence completion that humans solve easily.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-05-19T23:57:23Z","title":"HellaSwag: Can a Machine Really Finish Your Sentence?"},"references":{"count":19,"internal_anchors":2,"resolved_work":19,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Yonatan Belinkov and Yonatan Bisk. 2018. Synthetic and natural noise both break neural machine translation. In ICLR. ICLR","work_id":"2f87ef9e-973a-4f33-93e7-a79e927ca32e","year":2018},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"Qian Chen, Xiaodan Zhu, Zhen-Hua Ling, Si Wei, Hui Jiang, and Diana Inkpen. 2017. Enhanced lstm for natural language inference. In Proceedings of the 55th Annual Meeting of the Association for Computa","work_id":"8bd6683b-0e73-4ec0-bb86-391a37e265de","year":2017},{"cited_arxiv_id":"1810.04805","doi":"","is_internal_anchor":true,"ref_index":3,"title":"BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding","work_id":"ed240a10-5b19-406c-baa5-30803f465785","year":2018},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"Max Glockner, Vered Shwartz, and Yoav Goldberg. 2018. Breaking nli systems with sentences that require simple lexical inferences. In Proceedings of the 56th Annual Meeting of the Association for Compu","work_id":"277326a1-be96-4bb7-aab1-64021dbd9013","year":2018},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"Jonathan Gordon and Benjamin Van Durme. 2013. Reporting bias and knowledge acquisition. In Proceedings of the 2013 workshop on Automated knowledge base construction, pages 25--30. ACM","work_id":"1002e502-b170-4aca-bbbd-64a081e2f89b","year":2013}],"snapshot_sha256":"fc7576631fb4f5e083bb1553be3225ef4ffe26fe3c0026780127e85c0555c0e6"},"source":{"id":"1905.07830","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-11T02:51:33.151353Z","id":"4928dd65-3589-403a-bee7-d0205863269c","model_set":{"reader":"grok-4.3"},"one_line_summary":"HellaSwag dataset shows state-of-the-art models fail commonsense inference tasks that humans solve easily, built via adversarial filtering of distractors.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"HellaSwag shows state-of-the-art models still fail at commonsense sentence completion that humans solve easily.","strongest_claim":"Though its questions are trivial for humans (>95% accuracy), state-of-the-art models struggle (<48%).","weakest_assumption":"That adversarial filtering produces examples requiring genuine commonsense reasoning rather than merely exploiting specific weaknesses or distributional artifacts in the models used during filtering."}},"verdict_id":"4928dd65-3589-403a-bee7-d0205863269c"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:723630924116b90ce472fef958de768c50e42ebc083b24c778c4a7f47a409a85","target":"record","created_at":"2026-07-04T23:38:28Z","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":"927c4889782be6b6bd5c205399e62f8e4869cac9638c87bfc36eb6830589af43","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-05-19T23:57:23Z","title_canon_sha256":"e9e53d4b2c8e24587b3aa33ee5bfedbfafdf21af4617107e422e70fe02959910"},"schema_version":"1.0","source":{"id":"1905.07830","kind":"arxiv","version":1}},"canonical_sha256":"eb74dd9ae7054c9d8d0024420d107fdf956cf16e993c09c33a7d78713303065d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"eb74dd9ae7054c9d8d0024420d107fdf956cf16e993c09c33a7d78713303065d","first_computed_at":"2026-07-04T23:38:28.263807Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-04T23:38:28.263807Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8LYRE/EXerMYSN8Ff4AGCJDuWRzSl+GTvNhKaTd4lUOfx5Lc1RhWTOAgblWdvXrPdx3cV18/IrKLUmr6Af0YAw==","signature_status":"signed_v1","signed_at":"2026-07-04T23:38:28.264304Z","signed_message":"canonical_sha256_bytes"},"source_id":"1905.07830","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:723630924116b90ce472fef958de768c50e42ebc083b24c778c4a7f47a409a85","sha256:625a6428af8057ed1f663da335021806c87febdda38467521288fc7092efdceb"],"state_sha256":"780e79871f35bad9a745b27cda1f23bea2839d69de2bd647ce6a93b663c10063"}