{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:ITD6D27AS6HHGX6JHPINR6IHZF","short_pith_number":"pith:ITD6D27A","schema_version":"1.0","canonical_sha256":"44c7e1ebe0978e735fc93bd0d8f907c9626ba498f45c52a0f50c499abe3b05b1","source":{"kind":"arxiv","id":"1909.12434","version":2},"attestation_state":"computed","paper":{"title":"Learning the Difference that Makes a Difference with Counterfactually-Augmented Data","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","stat.ML"],"primary_cat":"cs.CL","authors_text":"Divyansh Kaushik, Eduard Hovy, Zachary C. Lipton","submitted_at":"2019-09-26T23:25:25Z","abstract_excerpt":"Despite alarm over the reliance of machine learning systems on so-called spurious patterns, the term lacks coherent meaning in standard statistical frameworks. However, the language of causality offers clarity: spurious associations are due to confounding (e.g., a common cause), but not direct or indirect causal effects. In this paper, we focus on natural language processing, introducing methods and resources for training models less sensitive to spurious patterns. Given documents and their initial labels, we task humans with revising each document so that it (i) accords with a counterfactual "},"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":"1909.12434","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-09-26T23:25:25Z","cross_cats_sorted":["cs.AI","cs.LG","stat.ML"],"title_canon_sha256":"36be252f2f3cdcf99713fb253811350a665a61757ac7332acb92b47b71b661a8","abstract_canon_sha256":"5f23939a328745f610c9c174d99e26f43eed5e2a721757ab7bb6c25a51582c3a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:40:58.689559Z","signature_b64":"xoCAbj3zQHMEeGKIrjji6Pr/B4QyMlD23x+OVs2wp+W7bxEwbtMYn4H7BGVAogkOiZOJlcSZLv99WUbcJoQHAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"44c7e1ebe0978e735fc93bd0d8f907c9626ba498f45c52a0f50c499abe3b05b1","last_reissued_at":"2026-07-05T00:40:58.689136Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:40:58.689136Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning the Difference that Makes a Difference with Counterfactually-Augmented Data","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","stat.ML"],"primary_cat":"cs.CL","authors_text":"Divyansh Kaushik, Eduard Hovy, Zachary C. Lipton","submitted_at":"2019-09-26T23:25:25Z","abstract_excerpt":"Despite alarm over the reliance of machine learning systems on so-called spurious patterns, the term lacks coherent meaning in standard statistical frameworks. However, the language of causality offers clarity: spurious associations are due to confounding (e.g., a common cause), but not direct or indirect causal effects. In this paper, we focus on natural language processing, introducing methods and resources for training models less sensitive to spurious patterns. Given documents and their initial labels, we task humans with revising each document so that it (i) accords with a counterfactual "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1909.12434","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/1909.12434/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":"1909.12434","created_at":"2026-07-05T00:40:58.689192+00:00"},{"alias_kind":"arxiv_version","alias_value":"1909.12434v2","created_at":"2026-07-05T00:40:58.689192+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1909.12434","created_at":"2026-07-05T00:40:58.689192+00:00"},{"alias_kind":"pith_short_12","alias_value":"ITD6D27AS6HH","created_at":"2026-07-05T00:40:58.689192+00:00"},{"alias_kind":"pith_short_16","alias_value":"ITD6D27AS6HHGX6J","created_at":"2026-07-05T00:40:58.689192+00:00"},{"alias_kind":"pith_short_8","alias_value":"ITD6D27A","created_at":"2026-07-05T00:40:58.689192+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.07865","citing_title":"Instrumented data for causal scientific machine learning","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06443","citing_title":"Revising Context, Shifting Simulated Stance: Auditing LLM-Based Stance Simulation in Online Discussions","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2008.02275","citing_title":"Aligning AI With Shared Human Values","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09716","citing_title":"Medical Model Synthesis Architectures: A Case Study","ref_index":183,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14325","citing_title":"Faithfulness Serum: Mitigating the Faithfulness Gap in Textual Explanations of LLM Decisions via Attribution Guidance","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09537","citing_title":"Case-Grounded Evidence Verification: A Framework for Constructing Evidence-Sensitive Supervision","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2604.16752","citing_title":"Don't Start What You Can't Finish: A Counterfactual Audit of Support-State Triage in LLM Agents","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ITD6D27AS6HHGX6JHPINR6IHZF","json":"https://pith.science/pith/ITD6D27AS6HHGX6JHPINR6IHZF.json","graph_json":"https://pith.science/api/pith-number/ITD6D27AS6HHGX6JHPINR6IHZF/graph.json","events_json":"https://pith.science/api/pith-number/ITD6D27AS6HHGX6JHPINR6IHZF/events.json","paper":"https://pith.science/paper/ITD6D27A"},"agent_actions":{"view_html":"https://pith.science/pith/ITD6D27AS6HHGX6JHPINR6IHZF","download_json":"https://pith.science/pith/ITD6D27AS6HHGX6JHPINR6IHZF.json","view_paper":"https://pith.science/paper/ITD6D27A","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1909.12434&json=true","fetch_graph":"https://pith.science/api/pith-number/ITD6D27AS6HHGX6JHPINR6IHZF/graph.json","fetch_events":"https://pith.science/api/pith-number/ITD6D27AS6HHGX6JHPINR6IHZF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ITD6D27AS6HHGX6JHPINR6IHZF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ITD6D27AS6HHGX6JHPINR6IHZF/action/storage_attestation","attest_author":"https://pith.science/pith/ITD6D27AS6HHGX6JHPINR6IHZF/action/author_attestation","sign_citation":"https://pith.science/pith/ITD6D27AS6HHGX6JHPINR6IHZF/action/citation_signature","submit_replication":"https://pith.science/pith/ITD6D27AS6HHGX6JHPINR6IHZF/action/replication_record"}},"created_at":"2026-07-05T00:40:58.689192+00:00","updated_at":"2026-07-05T00:40:58.689192+00:00"}