{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:5USESHWS7MQGYSZTI6QPNNAE7R","short_pith_number":"pith:5USESHWS","canonical_record":{"source":{"id":"2305.14324","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-23T17:54:57Z","cross_cats_sorted":[],"title_canon_sha256":"8e940bbb35048b9f7ca42b31dd901e0f956426d0277ed57b7279c82ce111e0d1","abstract_canon_sha256":"0920834939ff9f5e29c0bb0705dd61e5b7a1bb807294816e51d2f79713687f3e"},"schema_version":"1.0"},"canonical_sha256":"ed24491ed2fb206c4b3347a0f6b404fc5c26af5e846e84687c91ccfb881ee4ce","source":{"kind":"arxiv","id":"2305.14324","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.14324","created_at":"2026-07-05T07:01:41Z"},{"alias_kind":"arxiv_version","alias_value":"2305.14324v2","created_at":"2026-07-05T07:01:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.14324","created_at":"2026-07-05T07:01:41Z"},{"alias_kind":"pith_short_12","alias_value":"5USESHWS7MQG","created_at":"2026-07-05T07:01:41Z"},{"alias_kind":"pith_short_16","alias_value":"5USESHWS7MQGYSZT","created_at":"2026-07-05T07:01:41Z"},{"alias_kind":"pith_short_8","alias_value":"5USESHWS","created_at":"2026-07-05T07:01:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:5USESHWS7MQGYSZTI6QPNNAE7R","target":"record","payload":{"canonical_record":{"source":{"id":"2305.14324","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-23T17:54:57Z","cross_cats_sorted":[],"title_canon_sha256":"8e940bbb35048b9f7ca42b31dd901e0f956426d0277ed57b7279c82ce111e0d1","abstract_canon_sha256":"0920834939ff9f5e29c0bb0705dd61e5b7a1bb807294816e51d2f79713687f3e"},"schema_version":"1.0"},"canonical_sha256":"ed24491ed2fb206c4b3347a0f6b404fc5c26af5e846e84687c91ccfb881ee4ce","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:01:41.633325Z","signature_b64":"83K9qgm7HUpjTirEjlCNh1EYKfs6TDrJwgOt4PnTiymeG9x8ummVE8EGpXV8GsM2pch+gkYYUx+0PMTSvjiUDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ed24491ed2fb206c4b3347a0f6b404fc5c26af5e846e84687c91ccfb881ee4ce","last_reissued_at":"2026-07-05T07:01:41.632906Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:01:41.632906Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2305.14324","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":"2026-07-05T07:01:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iO5+3sg/SQBxsHRCUJsnDIJjANAdxdjj8EMyqO7RZQVkrii2o/Fo3soB69E8UIt/E52no3HxWzjd/gSrQ7HNDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T10:41:09.800384Z"},"content_sha256":"548415876cd9fccb6f92ef7d0e6405dbe15368085b6d308a12dcd2af4586e98f","schema_version":"1.0","event_id":"sha256:548415876cd9fccb6f92ef7d0e6405dbe15368085b6d308a12dcd2af4586e98f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:5USESHWS7MQGYSZTI6QPNNAE7R","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Ties Matter: Meta-Evaluating Modern Metrics with Pairwise Accuracy and Tie Calibration","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Daniel Deutsch, George Foster, Markus Freitag","submitted_at":"2023-05-23T17:54:57Z","abstract_excerpt":"Kendall's tau is frequently used to meta-evaluate how well machine translation (MT) evaluation metrics score individual translations. Its focus on pairwise score comparisons is intuitive but raises the question of how ties should be handled, a gray area that has motivated different variants in the literature. We demonstrate that, in settings like modern MT meta-evaluation, existing variants have weaknesses arising from their handling of ties, and in some situations can even be gamed. We propose instead to meta-evaluate metrics with a version of pairwise accuracy that gives metrics credit for c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.14324","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/2305.14324/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":"2026-07-05T07:01:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"64kflDkOhLCIKvMZJwC2B1QzBz+FbAyKnHziXX0QIMbjCV3i8kOWEo1NNfpjMZVvHHyLcp4ijBCnP+sHPvPMCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T10:41:09.800957Z"},"content_sha256":"814891d28f8c2b18478b3bc4d113d07f1e77e9fbc02e1b955d0a4c7175ed978a","schema_version":"1.0","event_id":"sha256:814891d28f8c2b18478b3bc4d113d07f1e77e9fbc02e1b955d0a4c7175ed978a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/5USESHWS7MQGYSZTI6QPNNAE7R/bundle.json","state_url":"https://pith.science/pith/5USESHWS7MQGYSZTI6QPNNAE7R/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/5USESHWS7MQGYSZTI6QPNNAE7R/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":"2026-08-05T10:41:09Z","links":{"resolver":"https://pith.science/pith/5USESHWS7MQGYSZTI6QPNNAE7R","bundle":"https://pith.science/pith/5USESHWS7MQGYSZTI6QPNNAE7R/bundle.json","state":"https://pith.science/pith/5USESHWS7MQGYSZTI6QPNNAE7R/state.json","well_known_bundle":"https://pith.science/.well-known/pith/5USESHWS7MQGYSZTI6QPNNAE7R/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:5USESHWS7MQGYSZTI6QPNNAE7R","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":"0920834939ff9f5e29c0bb0705dd61e5b7a1bb807294816e51d2f79713687f3e","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-23T17:54:57Z","title_canon_sha256":"8e940bbb35048b9f7ca42b31dd901e0f956426d0277ed57b7279c82ce111e0d1"},"schema_version":"1.0","source":{"id":"2305.14324","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.14324","created_at":"2026-07-05T07:01:41Z"},{"alias_kind":"arxiv_version","alias_value":"2305.14324v2","created_at":"2026-07-05T07:01:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.14324","created_at":"2026-07-05T07:01:41Z"},{"alias_kind":"pith_short_12","alias_value":"5USESHWS7MQG","created_at":"2026-07-05T07:01:41Z"},{"alias_kind":"pith_short_16","alias_value":"5USESHWS7MQGYSZT","created_at":"2026-07-05T07:01:41Z"},{"alias_kind":"pith_short_8","alias_value":"5USESHWS","created_at":"2026-07-05T07:01:41Z"}],"graph_snapshots":[{"event_id":"sha256:814891d28f8c2b18478b3bc4d113d07f1e77e9fbc02e1b955d0a4c7175ed978a","target":"graph","created_at":"2026-07-05T07:01:41Z","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/2305.14324/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Kendall's tau is frequently used to meta-evaluate how well machine translation (MT) evaluation metrics score individual translations. Its focus on pairwise score comparisons is intuitive but raises the question of how ties should be handled, a gray area that has motivated different variants in the literature. We demonstrate that, in settings like modern MT meta-evaluation, existing variants have weaknesses arising from their handling of ties, and in some situations can even be gamed. We propose instead to meta-evaluate metrics with a version of pairwise accuracy that gives metrics credit for c","authors_text":"Daniel Deutsch, George Foster, Markus Freitag","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-23T17:54:57Z","title":"Ties Matter: Meta-Evaluating Modern Metrics with Pairwise Accuracy and Tie Calibration"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.14324","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":"sha256:548415876cd9fccb6f92ef7d0e6405dbe15368085b6d308a12dcd2af4586e98f","target":"record","created_at":"2026-07-05T07:01:41Z","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":"0920834939ff9f5e29c0bb0705dd61e5b7a1bb807294816e51d2f79713687f3e","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-05-23T17:54:57Z","title_canon_sha256":"8e940bbb35048b9f7ca42b31dd901e0f956426d0277ed57b7279c82ce111e0d1"},"schema_version":"1.0","source":{"id":"2305.14324","kind":"arxiv","version":2}},"canonical_sha256":"ed24491ed2fb206c4b3347a0f6b404fc5c26af5e846e84687c91ccfb881ee4ce","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ed24491ed2fb206c4b3347a0f6b404fc5c26af5e846e84687c91ccfb881ee4ce","first_computed_at":"2026-07-05T07:01:41.632906Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:01:41.632906Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"83K9qgm7HUpjTirEjlCNh1EYKfs6TDrJwgOt4PnTiymeG9x8ummVE8EGpXV8GsM2pch+gkYYUx+0PMTSvjiUDw==","signature_status":"signed_v1","signed_at":"2026-07-05T07:01:41.633325Z","signed_message":"canonical_sha256_bytes"},"source_id":"2305.14324","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:548415876cd9fccb6f92ef7d0e6405dbe15368085b6d308a12dcd2af4586e98f","sha256:814891d28f8c2b18478b3bc4d113d07f1e77e9fbc02e1b955d0a4c7175ed978a"],"state_sha256":"042a1f9b78f0f314aa51c2b0a3a840d13ce107b4181acbec5464fe332f82c6b1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"X9Cef4S+pdV+kKRnqz70nCLqKLPbGH7/FoA1/s7c6oiL0op4GMWDY1vwhrC42h9aFrnzVfSj97mEIOEX0dLUBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T10:41:09.805098Z","bundle_sha256":"b101f2e3be1ac072545f63cada25bbe3f912df511b5bbe275b54d4d43a05bf71"}}