{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:ZBVHAP7S3UNPLAOPM4PMC7HPL2","short_pith_number":"pith:ZBVHAP7S","schema_version":"1.0","canonical_sha256":"c86a703ff2dd1af581cf671ec17cef5ea5d509e84366a18cfb0b472eeae00696","source":{"kind":"arxiv","id":"2202.07654","version":2},"attestation_state":"computed","paper":{"title":"Tomayto, Tomahto. Beyond Token-level Answer Equivalence for Question Answering Evaluation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Benjamin Boerschinger, Christian Buck, Jannis Bulian, Tal Schuster, Wojciech Gajewski","submitted_at":"2022-02-15T18:53:58Z","abstract_excerpt":"The predictions of question answering (QA)systems are typically evaluated against manually annotated finite sets of one or more answers. This leads to a coverage limitation that results in underestimating the true performance of systems, and is typically addressed by extending over exact match (EM) with pre-defined rules or with the token-level F1 measure. In this paper, we present the first systematic conceptual and data-driven analysis to examine the shortcomings of token-level equivalence measures.\n  To this end, we define the asymmetric notion of answer equivalence (AE), accepting answers "},"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":"2202.07654","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2022-02-15T18:53:58Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"f4c00de2777bf5d10ae88c21fe5ddce86487236795a59477891b495ab4807a20","abstract_canon_sha256":"19a9894713ddbb4341333791e6b7b0892d96e1282fab7375a02bd4ff66bb23e2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:10:36.640995Z","signature_b64":"AvOBN+61gsmVAZxRkVV4jYk+Ic0akTnC1btRZbAlLyacP16GcFV7BIM+RtPUzfT1VZnhH6D1xs9le08o93h9Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c86a703ff2dd1af581cf671ec17cef5ea5d509e84366a18cfb0b472eeae00696","last_reissued_at":"2026-07-05T05:10:36.640514Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:10:36.640514Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Tomayto, Tomahto. Beyond Token-level Answer Equivalence for Question Answering Evaluation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Benjamin Boerschinger, Christian Buck, Jannis Bulian, Tal Schuster, Wojciech Gajewski","submitted_at":"2022-02-15T18:53:58Z","abstract_excerpt":"The predictions of question answering (QA)systems are typically evaluated against manually annotated finite sets of one or more answers. This leads to a coverage limitation that results in underestimating the true performance of systems, and is typically addressed by extending over exact match (EM) with pre-defined rules or with the token-level F1 measure. In this paper, we present the first systematic conceptual and data-driven analysis to examine the shortcomings of token-level equivalence measures.\n  To this end, we define the asymmetric notion of answer equivalence (AE), accepting answers "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2202.07654","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/2202.07654/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":"2202.07654","created_at":"2026-07-05T05:10:36.640572+00:00"},{"alias_kind":"arxiv_version","alias_value":"2202.07654v2","created_at":"2026-07-05T05:10:36.640572+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2202.07654","created_at":"2026-07-05T05:10:36.640572+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZBVHAP7S3UNP","created_at":"2026-07-05T05:10:36.640572+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZBVHAP7S3UNPLAOP","created_at":"2026-07-05T05:10:36.640572+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZBVHAP7S","created_at":"2026-07-05T05:10:36.640572+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.12441","citing_title":"Describe Anything Model for Visual Question Answering on Text-rich Images","ref_index":7,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZBVHAP7S3UNPLAOPM4PMC7HPL2","json":"https://pith.science/pith/ZBVHAP7S3UNPLAOPM4PMC7HPL2.json","graph_json":"https://pith.science/api/pith-number/ZBVHAP7S3UNPLAOPM4PMC7HPL2/graph.json","events_json":"https://pith.science/api/pith-number/ZBVHAP7S3UNPLAOPM4PMC7HPL2/events.json","paper":"https://pith.science/paper/ZBVHAP7S"},"agent_actions":{"view_html":"https://pith.science/pith/ZBVHAP7S3UNPLAOPM4PMC7HPL2","download_json":"https://pith.science/pith/ZBVHAP7S3UNPLAOPM4PMC7HPL2.json","view_paper":"https://pith.science/paper/ZBVHAP7S","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2202.07654&json=true","fetch_graph":"https://pith.science/api/pith-number/ZBVHAP7S3UNPLAOPM4PMC7HPL2/graph.json","fetch_events":"https://pith.science/api/pith-number/ZBVHAP7S3UNPLAOPM4PMC7HPL2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZBVHAP7S3UNPLAOPM4PMC7HPL2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZBVHAP7S3UNPLAOPM4PMC7HPL2/action/storage_attestation","attest_author":"https://pith.science/pith/ZBVHAP7S3UNPLAOPM4PMC7HPL2/action/author_attestation","sign_citation":"https://pith.science/pith/ZBVHAP7S3UNPLAOPM4PMC7HPL2/action/citation_signature","submit_replication":"https://pith.science/pith/ZBVHAP7S3UNPLAOPM4PMC7HPL2/action/replication_record"}},"created_at":"2026-07-05T05:10:36.640572+00:00","updated_at":"2026-07-05T05:10:36.640572+00:00"}