{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:6NL6GISBCO5F4GNK4GQX3OS326","short_pith_number":"pith:6NL6GISB","schema_version":"1.0","canonical_sha256":"f357e3224113ba5e19aae1a17dba5bd7895309498df430613f8641613d168249","source":{"kind":"arxiv","id":"2004.14623","version":4},"attestation_state":"computed","paper":{"title":"Neural Natural Language Inference Models Partially Embed Theories of Lexical Entailment and Negation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Atticus Geiger, Christopher Potts, Kyle Richardson","submitted_at":"2020-04-30T07:53:20Z","abstract_excerpt":"We address whether neural models for Natural Language Inference (NLI) can learn the compositional interactions between lexical entailment and negation, using four methods: the behavioral evaluation methods of (1) challenge test sets and (2) systematic generalization tasks, and the structural evaluation methods of (3) probes and (4) interventions. To facilitate this holistic evaluation, we present Monotonicity NLI (MoNLI), a new naturalistic dataset focused on lexical entailment and negation. In our behavioral evaluations, we find that models trained on general-purpose NLI datasets fail systema"},"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":"2004.14623","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2020-04-30T07:53:20Z","cross_cats_sorted":[],"title_canon_sha256":"6065fd8de3ca1f3979f00cdef3904c6a77e0ae0007c10f4ec8a271f93aa8476f","abstract_canon_sha256":"50188e998666a78a9caebb7fed5481a9c283d01a8801a84e94811d000394129a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:53:19.978718Z","signature_b64":"tWdN6qLNlwhUPGyP90XsdhsSArPYAPuy0zO1VE1/u1zcJ7U1udurQFxbaMsxv1IsThVBWuuI49vsarhu88/qAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f357e3224113ba5e19aae1a17dba5bd7895309498df430613f8641613d168249","last_reissued_at":"2026-07-05T01:53:19.978372Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:53:19.978372Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Neural Natural Language Inference Models Partially Embed Theories of Lexical Entailment and Negation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Atticus Geiger, Christopher Potts, Kyle Richardson","submitted_at":"2020-04-30T07:53:20Z","abstract_excerpt":"We address whether neural models for Natural Language Inference (NLI) can learn the compositional interactions between lexical entailment and negation, using four methods: the behavioral evaluation methods of (1) challenge test sets and (2) systematic generalization tasks, and the structural evaluation methods of (3) probes and (4) interventions. To facilitate this holistic evaluation, we present Monotonicity NLI (MoNLI), a new naturalistic dataset focused on lexical entailment and negation. In our behavioral evaluations, we find that models trained on general-purpose NLI datasets fail systema"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.14623","kind":"arxiv","version":4},"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/2004.14623/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":"2004.14623","created_at":"2026-07-05T01:53:19.978431+00:00"},{"alias_kind":"arxiv_version","alias_value":"2004.14623v4","created_at":"2026-07-05T01:53:19.978431+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.14623","created_at":"2026-07-05T01:53:19.978431+00:00"},{"alias_kind":"pith_short_12","alias_value":"6NL6GISBCO5F","created_at":"2026-07-05T01:53:19.978431+00:00"},{"alias_kind":"pith_short_16","alias_value":"6NL6GISBCO5F4GNK","created_at":"2026-07-05T01:53:19.978431+00:00"},{"alias_kind":"pith_short_8","alias_value":"6NL6GISB","created_at":"2026-07-05T01:53:19.978431+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2404.15255","citing_title":"How to use and interpret activation patching","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2304.05969","citing_title":"Localizing Model Behavior with Path Patching","ref_index":39,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6NL6GISBCO5F4GNK4GQX3OS326","json":"https://pith.science/pith/6NL6GISBCO5F4GNK4GQX3OS326.json","graph_json":"https://pith.science/api/pith-number/6NL6GISBCO5F4GNK4GQX3OS326/graph.json","events_json":"https://pith.science/api/pith-number/6NL6GISBCO5F4GNK4GQX3OS326/events.json","paper":"https://pith.science/paper/6NL6GISB"},"agent_actions":{"view_html":"https://pith.science/pith/6NL6GISBCO5F4GNK4GQX3OS326","download_json":"https://pith.science/pith/6NL6GISBCO5F4GNK4GQX3OS326.json","view_paper":"https://pith.science/paper/6NL6GISB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2004.14623&json=true","fetch_graph":"https://pith.science/api/pith-number/6NL6GISBCO5F4GNK4GQX3OS326/graph.json","fetch_events":"https://pith.science/api/pith-number/6NL6GISBCO5F4GNK4GQX3OS326/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6NL6GISBCO5F4GNK4GQX3OS326/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6NL6GISBCO5F4GNK4GQX3OS326/action/storage_attestation","attest_author":"https://pith.science/pith/6NL6GISBCO5F4GNK4GQX3OS326/action/author_attestation","sign_citation":"https://pith.science/pith/6NL6GISBCO5F4GNK4GQX3OS326/action/citation_signature","submit_replication":"https://pith.science/pith/6NL6GISBCO5F4GNK4GQX3OS326/action/replication_record"}},"created_at":"2026-07-05T01:53:19.978431+00:00","updated_at":"2026-07-05T01:53:19.978431+00:00"}