{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:D2GSKB2WGYJOOCC7ADCB4BEQSD","short_pith_number":"pith:D2GSKB2W","canonical_record":{"source":{"id":"2410.16834","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-22T09:14:21Z","cross_cats_sorted":[],"title_canon_sha256":"f2a441ebd8f51e5702e7ed6f946970e0d8ecfb9e94cdd0fe51714376e13c7463","abstract_canon_sha256":"6a00258b8c3e5938c8c804d715364c90fc175f419111fa07edb1a8a474f29531"},"schema_version":"1.0"},"canonical_sha256":"1e8d2507563612e7085f00c41e049090cd20ec4c38fcadf2e21d6b595d5a6220","source":{"kind":"arxiv","id":"2410.16834","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.16834","created_at":"2026-07-05T10:05:13Z"},{"alias_kind":"arxiv_version","alias_value":"2410.16834v2","created_at":"2026-07-05T10:05:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.16834","created_at":"2026-07-05T10:05:13Z"},{"alias_kind":"pith_short_12","alias_value":"D2GSKB2WGYJO","created_at":"2026-07-05T10:05:13Z"},{"alias_kind":"pith_short_16","alias_value":"D2GSKB2WGYJOOCC7","created_at":"2026-07-05T10:05:13Z"},{"alias_kind":"pith_short_8","alias_value":"D2GSKB2W","created_at":"2026-07-05T10:05:13Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:D2GSKB2WGYJOOCC7ADCB4BEQSD","target":"record","payload":{"canonical_record":{"source":{"id":"2410.16834","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-22T09:14:21Z","cross_cats_sorted":[],"title_canon_sha256":"f2a441ebd8f51e5702e7ed6f946970e0d8ecfb9e94cdd0fe51714376e13c7463","abstract_canon_sha256":"6a00258b8c3e5938c8c804d715364c90fc175f419111fa07edb1a8a474f29531"},"schema_version":"1.0"},"canonical_sha256":"1e8d2507563612e7085f00c41e049090cd20ec4c38fcadf2e21d6b595d5a6220","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:05:13.847823Z","signature_b64":"/slvKVoFzCwMGfqYSK0wMatSZPdq+uodebLTCiHqC+6g+uBsbR+VU9WjBFjU9/hh1g2M+XVrnzVEmZFW8V6CAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1e8d2507563612e7085f00c41e049090cd20ec4c38fcadf2e21d6b595d5a6220","last_reissued_at":"2026-07-05T10:05:13.847277Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:05:13.847277Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.16834","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":"2026-07-05T10:05:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/hwx/VDYhFZ2Q0qpShWZLhOaeDBxGz2fYJYOAIGAM+ccSZsY8jkcL8+sQqVKuNRfX+9ep2BCQh3xBeHBaOjqDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T10:37:48.260033Z"},"content_sha256":"5f98d90f80275f8c8255f59b4da36e991eae0bea71e3f73e1e210e25fe1f3f06","schema_version":"1.0","event_id":"sha256:5f98d90f80275f8c8255f59b4da36e991eae0bea71e3f73e1e210e25fe1f3f06"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:D2GSKB2WGYJOOCC7ADCB4BEQSD","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Analyzing and Evaluating Correlation Measures in NLG Meta-Evaluation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Li Lin, Mingqi Gao, Xiaojun Wan, Xinyu Hu","submitted_at":"2024-10-22T09:14:21Z","abstract_excerpt":"The correlation between NLG automatic evaluation metrics and human evaluation is often regarded as a critical criterion for assessing the capability of an evaluation metric. However, different grouping methods and correlation coefficients result in various types of correlation measures used in meta-evaluation. In specific evaluation scenarios, prior work often directly follows conventional measure settings, but the characteristics and differences between these measures have not gotten sufficient attention. Therefore, this paper analyzes 12 common correlation measures using a large amount of re"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.16834","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/2410.16834/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":"2026-07-05T10:05:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"h1Hn0I/W2u8UmIaXoeDgys/B4BoW1FgFGw63elOYb0A0BXKKkLsV+fZovQ0h2ZYjGMCgdm1ChfVJqaahypL5BA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T10:37:48.260540Z"},"content_sha256":"78bc90e9d92b742b47719501515a79f7effe06aa8732b021a2842da7e9d0eae6","schema_version":"1.0","event_id":"sha256:78bc90e9d92b742b47719501515a79f7effe06aa8732b021a2842da7e9d0eae6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/D2GSKB2WGYJOOCC7ADCB4BEQSD/bundle.json","state_url":"https://pith.science/pith/D2GSKB2WGYJOOCC7ADCB4BEQSD/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/D2GSKB2WGYJOOCC7ADCB4BEQSD/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":"2026-08-04T10:37:48Z","links":{"resolver":"https://pith.science/pith/D2GSKB2WGYJOOCC7ADCB4BEQSD","bundle":"https://pith.science/pith/D2GSKB2WGYJOOCC7ADCB4BEQSD/bundle.json","state":"https://pith.science/pith/D2GSKB2WGYJOOCC7ADCB4BEQSD/state.json","well_known_bundle":"https://pith.science/.well-known/pith/D2GSKB2WGYJOOCC7ADCB4BEQSD/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:D2GSKB2WGYJOOCC7ADCB4BEQSD","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":"6a00258b8c3e5938c8c804d715364c90fc175f419111fa07edb1a8a474f29531","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-22T09:14:21Z","title_canon_sha256":"f2a441ebd8f51e5702e7ed6f946970e0d8ecfb9e94cdd0fe51714376e13c7463"},"schema_version":"1.0","source":{"id":"2410.16834","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.16834","created_at":"2026-07-05T10:05:13Z"},{"alias_kind":"arxiv_version","alias_value":"2410.16834v2","created_at":"2026-07-05T10:05:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.16834","created_at":"2026-07-05T10:05:13Z"},{"alias_kind":"pith_short_12","alias_value":"D2GSKB2WGYJO","created_at":"2026-07-05T10:05:13Z"},{"alias_kind":"pith_short_16","alias_value":"D2GSKB2WGYJOOCC7","created_at":"2026-07-05T10:05:13Z"},{"alias_kind":"pith_short_8","alias_value":"D2GSKB2W","created_at":"2026-07-05T10:05:13Z"}],"graph_snapshots":[{"event_id":"sha256:78bc90e9d92b742b47719501515a79f7effe06aa8732b021a2842da7e9d0eae6","target":"graph","created_at":"2026-07-05T10:05:13Z","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/2410.16834/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The correlation between NLG automatic evaluation metrics and human evaluation is often regarded as a critical criterion for assessing the capability of an evaluation metric. However, different grouping methods and correlation coefficients result in various types of correlation measures used in meta-evaluation. In specific evaluation scenarios, prior work often directly follows conventional measure settings, but the characteristics and differences between these measures have not gotten sufficient attention. Therefore, this paper analyzes 12 common correlation measures using a large amount of re","authors_text":"Li Lin, Mingqi Gao, Xiaojun Wan, Xinyu Hu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-22T09:14:21Z","title":"Analyzing and Evaluating Correlation Measures in NLG Meta-Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.16834","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":"sha256:5f98d90f80275f8c8255f59b4da36e991eae0bea71e3f73e1e210e25fe1f3f06","target":"record","created_at":"2026-07-05T10:05:13Z","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":"6a00258b8c3e5938c8c804d715364c90fc175f419111fa07edb1a8a474f29531","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-22T09:14:21Z","title_canon_sha256":"f2a441ebd8f51e5702e7ed6f946970e0d8ecfb9e94cdd0fe51714376e13c7463"},"schema_version":"1.0","source":{"id":"2410.16834","kind":"arxiv","version":2}},"canonical_sha256":"1e8d2507563612e7085f00c41e049090cd20ec4c38fcadf2e21d6b595d5a6220","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1e8d2507563612e7085f00c41e049090cd20ec4c38fcadf2e21d6b595d5a6220","first_computed_at":"2026-07-05T10:05:13.847277Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:05:13.847277Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"/slvKVoFzCwMGfqYSK0wMatSZPdq+uodebLTCiHqC+6g+uBsbR+VU9WjBFjU9/hh1g2M+XVrnzVEmZFW8V6CAg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:05:13.847823Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.16834","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5f98d90f80275f8c8255f59b4da36e991eae0bea71e3f73e1e210e25fe1f3f06","sha256:78bc90e9d92b742b47719501515a79f7effe06aa8732b021a2842da7e9d0eae6"],"state_sha256":"eaebf3f7f899fba17c9a2ef59a4166e2f61fbfe165a5bc9fb031c073ae86f18f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"acffSx3gWn9V1KuU3X7VkWx/L2XfLGCoixRHAArw91qtzpe82W+s9f7Ab5A+JjvQdi4gUPsi2ZXzdKboin96Bg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T10:37:48.264338Z","bundle_sha256":"232f30ab5fdbb8d48b293b07442f9c478e23966f1739ac6e24c89d65c7964369"}}