{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:65O67VBP7I33ESJVKPX5HSF3C6","short_pith_number":"pith:65O67VBP","schema_version":"1.0","canonical_sha256":"f75defd42ffa37b2493553efd3c8bb17a035f26e143dc7ccbb0387fd994eb2e8","source":{"kind":"arxiv","id":"2010.03532","version":2},"attestation_state":"computed","paper":{"title":"What Can We Learn from Collective Human Opinions on Natural Language Inference Data?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Mohit Bansal, Xiang Zhou, Yixin Nie","submitted_at":"2020-10-07T17:26:06Z","abstract_excerpt":"Despite the subjective nature of many NLP tasks, most NLU evaluations have focused on using the majority label with presumably high agreement as the ground truth. Less attention has been paid to the distribution of human opinions. We collect ChaosNLI, a dataset with a total of 464,500 annotations to study Collective HumAn OpinionS in oft-used NLI evaluation sets. This dataset is created by collecting 100 annotations per example for 3,113 examples in SNLI and MNLI and 1,532 examples in Abductive-NLI. Analysis reveals that: (1) high human disagreement exists in a noticeable amount of examples in"},"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":"2010.03532","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-10-07T17:26:06Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"fd0afffdfaf43acdc6ca1302aa661411bac15f28843796ae129abd994754560a","abstract_canon_sha256":"e217f833de65f7a83a5a4e89f34e4b625e8fcaf2af6a00a5cba780894af1b265"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:41:34.109566Z","signature_b64":"JzJt8CFW0/GUGnLloY1fAHNpA0QSgnunqFNVPHOamCQNeMAMGBuml09es6VMTVVh/SbjlAeu5FHLUaajzq0DDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f75defd42ffa37b2493553efd3c8bb17a035f26e143dc7ccbb0387fd994eb2e8","last_reissued_at":"2026-07-05T01:41:34.109186Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:41:34.109186Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"What Can We Learn from Collective Human Opinions on Natural Language Inference Data?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Mohit Bansal, Xiang Zhou, Yixin Nie","submitted_at":"2020-10-07T17:26:06Z","abstract_excerpt":"Despite the subjective nature of many NLP tasks, most NLU evaluations have focused on using the majority label with presumably high agreement as the ground truth. Less attention has been paid to the distribution of human opinions. We collect ChaosNLI, a dataset with a total of 464,500 annotations to study Collective HumAn OpinionS in oft-used NLI evaluation sets. This dataset is created by collecting 100 annotations per example for 3,113 examples in SNLI and MNLI and 1,532 examples in Abductive-NLI. Analysis reveals that: (1) high human disagreement exists in a noticeable amount of examples in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.03532","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/2010.03532/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":"2010.03532","created_at":"2026-07-05T01:41:34.109239+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.03532v2","created_at":"2026-07-05T01:41:34.109239+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.03532","created_at":"2026-07-05T01:41:34.109239+00:00"},{"alias_kind":"pith_short_12","alias_value":"65O67VBP7I33","created_at":"2026-07-05T01:41:34.109239+00:00"},{"alias_kind":"pith_short_16","alias_value":"65O67VBP7I33ESJV","created_at":"2026-07-05T01:41:34.109239+00:00"},{"alias_kind":"pith_short_8","alias_value":"65O67VBP","created_at":"2026-07-05T01:41:34.109239+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/65O67VBP7I33ESJVKPX5HSF3C6","json":"https://pith.science/pith/65O67VBP7I33ESJVKPX5HSF3C6.json","graph_json":"https://pith.science/api/pith-number/65O67VBP7I33ESJVKPX5HSF3C6/graph.json","events_json":"https://pith.science/api/pith-number/65O67VBP7I33ESJVKPX5HSF3C6/events.json","paper":"https://pith.science/paper/65O67VBP"},"agent_actions":{"view_html":"https://pith.science/pith/65O67VBP7I33ESJVKPX5HSF3C6","download_json":"https://pith.science/pith/65O67VBP7I33ESJVKPX5HSF3C6.json","view_paper":"https://pith.science/paper/65O67VBP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.03532&json=true","fetch_graph":"https://pith.science/api/pith-number/65O67VBP7I33ESJVKPX5HSF3C6/graph.json","fetch_events":"https://pith.science/api/pith-number/65O67VBP7I33ESJVKPX5HSF3C6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/65O67VBP7I33ESJVKPX5HSF3C6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/65O67VBP7I33ESJVKPX5HSF3C6/action/storage_attestation","attest_author":"https://pith.science/pith/65O67VBP7I33ESJVKPX5HSF3C6/action/author_attestation","sign_citation":"https://pith.science/pith/65O67VBP7I33ESJVKPX5HSF3C6/action/citation_signature","submit_replication":"https://pith.science/pith/65O67VBP7I33ESJVKPX5HSF3C6/action/replication_record"}},"created_at":"2026-07-05T01:41:34.109239+00:00","updated_at":"2026-07-05T01:41:34.109239+00:00"}