{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:F57ELTNXHWBEOI26YSB7S4TYBP","short_pith_number":"pith:F57ELTNX","canonical_record":{"source":{"id":"2305.14991","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T10:26:13Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"bf96cfc2e883995911ac4f1726901ed86ce0c801e3a57eb27c12866ba39097a6","abstract_canon_sha256":"3d9c4808f2a4973bbc79af427852b2322cdf83e04c2aa7f16224379eac7018ba"},"schema_version":"1.0"},"canonical_sha256":"2f7e45cdb73d8247235ec483f972780bf1a6283bffeedd59390d3c172d10fedb","source":{"kind":"arxiv","id":"2305.14991","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.14991","created_at":"2026-07-05T07:18:10Z"},{"alias_kind":"arxiv_version","alias_value":"2305.14991v2","created_at":"2026-07-05T07:18:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.14991","created_at":"2026-07-05T07:18:10Z"},{"alias_kind":"pith_short_12","alias_value":"F57ELTNXHWBE","created_at":"2026-07-05T07:18:10Z"},{"alias_kind":"pith_short_16","alias_value":"F57ELTNXHWBEOI26","created_at":"2026-07-05T07:18:10Z"},{"alias_kind":"pith_short_8","alias_value":"F57ELTNX","created_at":"2026-07-05T07:18:10Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:F57ELTNXHWBEOI26YSB7S4TYBP","target":"record","payload":{"canonical_record":{"source":{"id":"2305.14991","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T10:26:13Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"bf96cfc2e883995911ac4f1726901ed86ce0c801e3a57eb27c12866ba39097a6","abstract_canon_sha256":"3d9c4808f2a4973bbc79af427852b2322cdf83e04c2aa7f16224379eac7018ba"},"schema_version":"1.0"},"canonical_sha256":"2f7e45cdb73d8247235ec483f972780bf1a6283bffeedd59390d3c172d10fedb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:18:10.239663Z","signature_b64":"Y/UTnXKMVcdoDM/l5vDfg/zTTfhOYH4GJZRWiFHKIbviyOVS5cB7kQzciaxRqM4rf15IThbLB5ir9/UUDF1aCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2f7e45cdb73d8247235ec483f972780bf1a6283bffeedd59390d3c172d10fedb","last_reissued_at":"2026-07-05T07:18:10.239174Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:18:10.239174Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2305.14991","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":"2026-07-05T07:18:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+OvRShfBGmr7OtSqZ9m6fN4bBA0g6OnR85jelhavTqv5RVr7RvW/HSXW//8qm8f7iAH2z7Veh1/n7EWDgP/VCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T11:12:49.573593Z"},"content_sha256":"5d3c4307bc6492fad0ba09eb313cb7a4057db6cf16a2346ba84e2625bfec7973","schema_version":"1.0","event_id":"sha256:5d3c4307bc6492fad0ba09eb313cb7a4057db6cf16a2346ba84e2625bfec7973"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:F57ELTNXHWBEOI26YSB7S4TYBP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"MuLER: Detailed and Scalable Reference-based Evaluation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Gal Patel, Leshem Choshen, Omri Abend, Taelin Karidi","submitted_at":"2023-05-24T10:26:13Z","abstract_excerpt":"We propose a novel methodology (namely, MuLER) that transforms any reference-based evaluation metric for text generation, such as machine translation (MT) into a fine-grained analysis tool. Given a system and a metric, MuLER quantifies how much the chosen metric penalizes specific error types (e.g., errors in translating names of locations). MuLER thus enables a detailed error analysis which can lead to targeted improvement efforts for specific phenomena. We perform experiments in both synthetic and naturalistic settings to support MuLER's validity and showcase its usability in MT evaluation, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.14991","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/2305.14991/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":"2026-07-05T07:18:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"T/8JDs6Ot05FuFqkZrDrtvxdKdRGo9o5zTMhrpy9oWl3hQl78LZE7ILvJxL6Y76pr//ValxAcxVkK+xOm5RYAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T11:12:49.574109Z"},"content_sha256":"387af666cf173caf0f12db4ff050f45672bfab3df455eb9e918261c805def26c","schema_version":"1.0","event_id":"sha256:387af666cf173caf0f12db4ff050f45672bfab3df455eb9e918261c805def26c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/F57ELTNXHWBEOI26YSB7S4TYBP/bundle.json","state_url":"https://pith.science/pith/F57ELTNXHWBEOI26YSB7S4TYBP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/F57ELTNXHWBEOI26YSB7S4TYBP/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":"2026-08-05T11:12:49Z","links":{"resolver":"https://pith.science/pith/F57ELTNXHWBEOI26YSB7S4TYBP","bundle":"https://pith.science/pith/F57ELTNXHWBEOI26YSB7S4TYBP/bundle.json","state":"https://pith.science/pith/F57ELTNXHWBEOI26YSB7S4TYBP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/F57ELTNXHWBEOI26YSB7S4TYBP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:F57ELTNXHWBEOI26YSB7S4TYBP","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":"3d9c4808f2a4973bbc79af427852b2322cdf83e04c2aa7f16224379eac7018ba","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T10:26:13Z","title_canon_sha256":"bf96cfc2e883995911ac4f1726901ed86ce0c801e3a57eb27c12866ba39097a6"},"schema_version":"1.0","source":{"id":"2305.14991","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2305.14991","created_at":"2026-07-05T07:18:10Z"},{"alias_kind":"arxiv_version","alias_value":"2305.14991v2","created_at":"2026-07-05T07:18:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.14991","created_at":"2026-07-05T07:18:10Z"},{"alias_kind":"pith_short_12","alias_value":"F57ELTNXHWBE","created_at":"2026-07-05T07:18:10Z"},{"alias_kind":"pith_short_16","alias_value":"F57ELTNXHWBEOI26","created_at":"2026-07-05T07:18:10Z"},{"alias_kind":"pith_short_8","alias_value":"F57ELTNX","created_at":"2026-07-05T07:18:10Z"}],"graph_snapshots":[{"event_id":"sha256:387af666cf173caf0f12db4ff050f45672bfab3df455eb9e918261c805def26c","target":"graph","created_at":"2026-07-05T07:18:10Z","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/2305.14991/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We propose a novel methodology (namely, MuLER) that transforms any reference-based evaluation metric for text generation, such as machine translation (MT) into a fine-grained analysis tool. Given a system and a metric, MuLER quantifies how much the chosen metric penalizes specific error types (e.g., errors in translating names of locations). MuLER thus enables a detailed error analysis which can lead to targeted improvement efforts for specific phenomena. We perform experiments in both synthetic and naturalistic settings to support MuLER's validity and showcase its usability in MT evaluation, ","authors_text":"Gal Patel, Leshem Choshen, Omri Abend, Taelin Karidi","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T10:26:13Z","title":"MuLER: Detailed and Scalable Reference-based Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.14991","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":"sha256:5d3c4307bc6492fad0ba09eb313cb7a4057db6cf16a2346ba84e2625bfec7973","target":"record","created_at":"2026-07-05T07:18:10Z","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":"3d9c4808f2a4973bbc79af427852b2322cdf83e04c2aa7f16224379eac7018ba","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-05-24T10:26:13Z","title_canon_sha256":"bf96cfc2e883995911ac4f1726901ed86ce0c801e3a57eb27c12866ba39097a6"},"schema_version":"1.0","source":{"id":"2305.14991","kind":"arxiv","version":2}},"canonical_sha256":"2f7e45cdb73d8247235ec483f972780bf1a6283bffeedd59390d3c172d10fedb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2f7e45cdb73d8247235ec483f972780bf1a6283bffeedd59390d3c172d10fedb","first_computed_at":"2026-07-05T07:18:10.239174Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:18:10.239174Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Y/UTnXKMVcdoDM/l5vDfg/zTTfhOYH4GJZRWiFHKIbviyOVS5cB7kQzciaxRqM4rf15IThbLB5ir9/UUDF1aCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T07:18:10.239663Z","signed_message":"canonical_sha256_bytes"},"source_id":"2305.14991","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5d3c4307bc6492fad0ba09eb313cb7a4057db6cf16a2346ba84e2625bfec7973","sha256:387af666cf173caf0f12db4ff050f45672bfab3df455eb9e918261c805def26c"],"state_sha256":"2a631bf83b12bcfb9370f3eaf349c83db2d4ccd4d3527513435c73eb14cc4919"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zAbW7qpsGHYsaZFwF32rq4iNGh6JV6eHstlLTU0pcS3WwlN/+4FlIrusdMgvlGLEPVA9zxh61qjWQa3wqGbSBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T11:12:49.579557Z","bundle_sha256":"17901d79a1cbe6b5467b3c99e295afe4428b7117a577e7b559200081fb0bdffd"}}