{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:YETMAQ2ZMUMLQGHKNXKRGRNWNU","short_pith_number":"pith:YETMAQ2Z","canonical_record":{"source":{"id":"2502.09416","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-13T15:39:07Z","cross_cats_sorted":[],"title_canon_sha256":"fb5d1b8baf66cd8ede242f154c9b1a1944ae4cd694d489d82db5c1a6648cabc5","abstract_canon_sha256":"e91adc8be1d77491b371517889f76f4cc74f881a25a5d3aafdfea9e6361656bc"},"schema_version":"1.0"},"canonical_sha256":"c126c043596518b818ea6dd51345b66d26a07d552a86cc1d6f2d478375652e8a","source":{"kind":"arxiv","id":"2502.09416","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.09416","created_at":"2026-07-05T11:14:53Z"},{"alias_kind":"arxiv_version","alias_value":"2502.09416v2","created_at":"2026-07-05T11:14:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.09416","created_at":"2026-07-05T11:14:53Z"},{"alias_kind":"pith_short_12","alias_value":"YETMAQ2ZMUML","created_at":"2026-07-05T11:14:53Z"},{"alias_kind":"pith_short_16","alias_value":"YETMAQ2ZMUMLQGHK","created_at":"2026-07-05T11:14:53Z"},{"alias_kind":"pith_short_8","alias_value":"YETMAQ2Z","created_at":"2026-07-05T11:14:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:YETMAQ2ZMUMLQGHKNXKRGRNWNU","target":"record","payload":{"canonical_record":{"source":{"id":"2502.09416","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-13T15:39:07Z","cross_cats_sorted":[],"title_canon_sha256":"fb5d1b8baf66cd8ede242f154c9b1a1944ae4cd694d489d82db5c1a6648cabc5","abstract_canon_sha256":"e91adc8be1d77491b371517889f76f4cc74f881a25a5d3aafdfea9e6361656bc"},"schema_version":"1.0"},"canonical_sha256":"c126c043596518b818ea6dd51345b66d26a07d552a86cc1d6f2d478375652e8a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:14:53.618252Z","signature_b64":"4grgIC0dB6UX6FSr4Rgkfulsg7k56z9CC7bYEkqLSG00B3tA9q0HjYE0dbq/mcxWesGpo0zJDtB+K9HmETOODg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c126c043596518b818ea6dd51345b66d26a07d552a86cc1d6f2d478375652e8a","last_reissued_at":"2026-07-05T11:14:53.617637Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:14:53.617637Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.09416","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:14:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PThEog+MpCrh2Gi1nEUehdvIUdzwtwYVUYNJ+uyiU10z2QBZsb7ndHgjLSLjoedEq6KwKddSi31WYxiPpkvRAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T03:43:44.837960Z"},"content_sha256":"5ecba0664eec97f05b62e024a9d59e2220a3b58eab6a6c853f182a3c40b4fccb","schema_version":"1.0","event_id":"sha256:5ecba0664eec97f05b62e024a9d59e2220a3b58eab6a6c853f182a3c40b4fccb"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:YETMAQ2ZMUMLQGHKNXKRGRNWNU","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Rethinking Evaluation Metrics for Grammatical Error Correction: Why Use a Different Evaluation Process than Human?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Takumi Goto, Taro Watanabe, Yusuke Sakai","submitted_at":"2025-02-13T15:39:07Z","abstract_excerpt":"One of the goals of automatic evaluation metrics in grammatical error correction (GEC) is to rank GEC systems such that it matches human preferences. However, current automatic evaluations are based on procedures that diverge from human evaluation. Specifically, human evaluation derives rankings by aggregating sentence-level relative evaluation results, e.g., pairwise comparisons, using a rating algorithm, whereas automatic evaluation averages sentence-level absolute scores to obtain corpus-level scores, which are then sorted to determine rankings. In this study, we propose an aggregation meth"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.09416","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/2502.09416/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:14:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nWmlxXt7DNbiEQ6OajYwgjriMhPGEt4jmn2fdsFXylcT7ig/DZdg/Tk+0oPxNxGJRufdfvYxD+SlCnj10FIbCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T03:43:44.838484Z"},"content_sha256":"bc8f1f51aae2e1070ed7213e110a1382d3325fe7866321fd3e5f6111a668ef2d","schema_version":"1.0","event_id":"sha256:bc8f1f51aae2e1070ed7213e110a1382d3325fe7866321fd3e5f6111a668ef2d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/YETMAQ2ZMUMLQGHKNXKRGRNWNU/bundle.json","state_url":"https://pith.science/pith/YETMAQ2ZMUMLQGHKNXKRGRNWNU/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/YETMAQ2ZMUMLQGHKNXKRGRNWNU/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-11T03:43:44Z","links":{"resolver":"https://pith.science/pith/YETMAQ2ZMUMLQGHKNXKRGRNWNU","bundle":"https://pith.science/pith/YETMAQ2ZMUMLQGHKNXKRGRNWNU/bundle.json","state":"https://pith.science/pith/YETMAQ2ZMUMLQGHKNXKRGRNWNU/state.json","well_known_bundle":"https://pith.science/.well-known/pith/YETMAQ2ZMUMLQGHKNXKRGRNWNU/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:YETMAQ2ZMUMLQGHKNXKRGRNWNU","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":"e91adc8be1d77491b371517889f76f4cc74f881a25a5d3aafdfea9e6361656bc","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-13T15:39:07Z","title_canon_sha256":"fb5d1b8baf66cd8ede242f154c9b1a1944ae4cd694d489d82db5c1a6648cabc5"},"schema_version":"1.0","source":{"id":"2502.09416","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.09416","created_at":"2026-07-05T11:14:53Z"},{"alias_kind":"arxiv_version","alias_value":"2502.09416v2","created_at":"2026-07-05T11:14:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.09416","created_at":"2026-07-05T11:14:53Z"},{"alias_kind":"pith_short_12","alias_value":"YETMAQ2ZMUML","created_at":"2026-07-05T11:14:53Z"},{"alias_kind":"pith_short_16","alias_value":"YETMAQ2ZMUMLQGHK","created_at":"2026-07-05T11:14:53Z"},{"alias_kind":"pith_short_8","alias_value":"YETMAQ2Z","created_at":"2026-07-05T11:14:53Z"}],"graph_snapshots":[{"event_id":"sha256:bc8f1f51aae2e1070ed7213e110a1382d3325fe7866321fd3e5f6111a668ef2d","target":"graph","created_at":"2026-07-05T11:14:53Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2502.09416/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"One of the goals of automatic evaluation metrics in grammatical error correction (GEC) is to rank GEC systems such that it matches human preferences. However, current automatic evaluations are based on procedures that diverge from human evaluation. Specifically, human evaluation derives rankings by aggregating sentence-level relative evaluation results, e.g., pairwise comparisons, using a rating algorithm, whereas automatic evaluation averages sentence-level absolute scores to obtain corpus-level scores, which are then sorted to determine rankings. In this study, we propose an aggregation meth","authors_text":"Takumi Goto, Taro Watanabe, Yusuke Sakai","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-13T15:39:07Z","title":"Rethinking Evaluation Metrics for Grammatical Error Correction: Why Use a Different Evaluation Process than Human?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.09416","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:5ecba0664eec97f05b62e024a9d59e2220a3b58eab6a6c853f182a3c40b4fccb","target":"record","created_at":"2026-07-05T11:14:53Z","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":"e91adc8be1d77491b371517889f76f4cc74f881a25a5d3aafdfea9e6361656bc","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-13T15:39:07Z","title_canon_sha256":"fb5d1b8baf66cd8ede242f154c9b1a1944ae4cd694d489d82db5c1a6648cabc5"},"schema_version":"1.0","source":{"id":"2502.09416","kind":"arxiv","version":2}},"canonical_sha256":"c126c043596518b818ea6dd51345b66d26a07d552a86cc1d6f2d478375652e8a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c126c043596518b818ea6dd51345b66d26a07d552a86cc1d6f2d478375652e8a","first_computed_at":"2026-07-05T11:14:53.617637Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:14:53.617637Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4grgIC0dB6UX6FSr4Rgkfulsg7k56z9CC7bYEkqLSG00B3tA9q0HjYE0dbq/mcxWesGpo0zJDtB+K9HmETOODg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:14:53.618252Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.09416","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5ecba0664eec97f05b62e024a9d59e2220a3b58eab6a6c853f182a3c40b4fccb","sha256:bc8f1f51aae2e1070ed7213e110a1382d3325fe7866321fd3e5f6111a668ef2d"],"state_sha256":"97a80f09cbdbc7b923e805176d82b9b3bcdd27717b25ec2b8e48a1d7c20619ec"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Yz86BsQQiZe6ZfE4DbTvUTLGgEZMtXkBbN0/tpjgzQhIPFDyjd6SGlVAciYtsXGbOKONWmWKJ2QaSdD+aliGDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-11T03:43:44.850187Z","bundle_sha256":"f8e68edcb22d33e3c07dd85ef557c67ba2ba2c0074bdba539b85c272eb312ad9"}}