{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:YTFSJT3YA232UQ4WXV6HI2S5HP","short_pith_number":"pith:YTFSJT3Y","canonical_record":{"source":{"id":"2410.00863","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-01T16:59:01Z","cross_cats_sorted":[],"title_canon_sha256":"f13b168ff17ad1b960c32e9ca6e5c94dea3f8baa71e3f75fb9d44d5464bd4733","abstract_canon_sha256":"d4c1722c9a3ed66b678aff471a40bb1251f325c7b87e3e5c0ae9e18a9a8e4e37"},"schema_version":"1.0"},"canonical_sha256":"c4cb24cf7806b7aa4396bd7c746a5d3bc6cf7a2b44f5b6b40fb85083f1db6a8b","source":{"kind":"arxiv","id":"2410.00863","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.00863","created_at":"2026-07-05T09:14:23Z"},{"alias_kind":"arxiv_version","alias_value":"2410.00863v1","created_at":"2026-07-05T09:14:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.00863","created_at":"2026-07-05T09:14:23Z"},{"alias_kind":"pith_short_12","alias_value":"YTFSJT3YA232","created_at":"2026-07-05T09:14:23Z"},{"alias_kind":"pith_short_16","alias_value":"YTFSJT3YA232UQ4W","created_at":"2026-07-05T09:14:23Z"},{"alias_kind":"pith_short_8","alias_value":"YTFSJT3Y","created_at":"2026-07-05T09:14:23Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:YTFSJT3YA232UQ4WXV6HI2S5HP","target":"record","payload":{"canonical_record":{"source":{"id":"2410.00863","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-01T16:59:01Z","cross_cats_sorted":[],"title_canon_sha256":"f13b168ff17ad1b960c32e9ca6e5c94dea3f8baa71e3f75fb9d44d5464bd4733","abstract_canon_sha256":"d4c1722c9a3ed66b678aff471a40bb1251f325c7b87e3e5c0ae9e18a9a8e4e37"},"schema_version":"1.0"},"canonical_sha256":"c4cb24cf7806b7aa4396bd7c746a5d3bc6cf7a2b44f5b6b40fb85083f1db6a8b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:14:23.880323Z","signature_b64":"aw+9QPHn4w/Lbcy/9jjIe8IiAqxeR6UtrCCA4G10ouxA4OaySEnzFRqiktWYpTkriIdOQcJd4shKQGE7czHPCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c4cb24cf7806b7aa4396bd7c746a5d3bc6cf7a2b44f5b6b40fb85083f1db6a8b","last_reissued_at":"2026-07-05T09:14:23.879818Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:14:23.879818Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.00863","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:14:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oDKoHL11WNA73jwtUKLbyhlHDuFyQSui7To+9HQhyuKYYnxiaNHQToaLdWp3MPU38r2J65pVdjskYqH0dnHYBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T13:27:06.763999Z"},"content_sha256":"6eede11272cee11c209aaa3440f60f87713ccab14a390aa4d59dc376162c8d70","schema_version":"1.0","event_id":"sha256:6eede11272cee11c209aaa3440f60f87713ccab14a390aa4d59dc376162c8d70"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:YTFSJT3YA232UQ4WXV6HI2S5HP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"On the Implications of Verbose LLM Outputs: A Case Study in Translation Evaluation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Colin Cherry, Eleftheria Briakou, Markus Freitag, Zhongtao Liu","submitted_at":"2024-10-01T16:59:01Z","abstract_excerpt":"This paper investigates the impact of verbose LLM translations on evaluation. We first demonstrate the prevalence of this behavior across several LLM outputs drawn from the WMT 2024 general shared task on machine translation. We then identify the primary triggers of verbosity, including safety, copyright concerns, and insufficient context in short input queries. Finally, we show that ignoring this behavior unfairly penalizes more verbose LLMs according to both automatic and human evaluations, highlighting the need to address this issue for more accurate future evaluations."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.00863","kind":"arxiv","version":1},"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/2410.00863/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":"2026-07-05T09:14:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0rLTb5r+iwmkStGfQOtW7PP7nNAboQTZ0TMojFRcBoUaxysDJGLKc8A/S0M7d2mUa5S39S0miy28eys784VJDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T13:27:06.764527Z"},"content_sha256":"8fc888225e820fb4da663ee4b1980e622bcfde6eb90d23f72d19904122949abc","schema_version":"1.0","event_id":"sha256:8fc888225e820fb4da663ee4b1980e622bcfde6eb90d23f72d19904122949abc"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/YTFSJT3YA232UQ4WXV6HI2S5HP/bundle.json","state_url":"https://pith.science/pith/YTFSJT3YA232UQ4WXV6HI2S5HP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/YTFSJT3YA232UQ4WXV6HI2S5HP/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":"2026-08-05T13:27:06Z","links":{"resolver":"https://pith.science/pith/YTFSJT3YA232UQ4WXV6HI2S5HP","bundle":"https://pith.science/pith/YTFSJT3YA232UQ4WXV6HI2S5HP/bundle.json","state":"https://pith.science/pith/YTFSJT3YA232UQ4WXV6HI2S5HP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/YTFSJT3YA232UQ4WXV6HI2S5HP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:YTFSJT3YA232UQ4WXV6HI2S5HP","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":"d4c1722c9a3ed66b678aff471a40bb1251f325c7b87e3e5c0ae9e18a9a8e4e37","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-01T16:59:01Z","title_canon_sha256":"f13b168ff17ad1b960c32e9ca6e5c94dea3f8baa71e3f75fb9d44d5464bd4733"},"schema_version":"1.0","source":{"id":"2410.00863","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.00863","created_at":"2026-07-05T09:14:23Z"},{"alias_kind":"arxiv_version","alias_value":"2410.00863v1","created_at":"2026-07-05T09:14:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.00863","created_at":"2026-07-05T09:14:23Z"},{"alias_kind":"pith_short_12","alias_value":"YTFSJT3YA232","created_at":"2026-07-05T09:14:23Z"},{"alias_kind":"pith_short_16","alias_value":"YTFSJT3YA232UQ4W","created_at":"2026-07-05T09:14:23Z"},{"alias_kind":"pith_short_8","alias_value":"YTFSJT3Y","created_at":"2026-07-05T09:14:23Z"}],"graph_snapshots":[{"event_id":"sha256:8fc888225e820fb4da663ee4b1980e622bcfde6eb90d23f72d19904122949abc","target":"graph","created_at":"2026-07-05T09:14:23Z","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/2410.00863/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper investigates the impact of verbose LLM translations on evaluation. We first demonstrate the prevalence of this behavior across several LLM outputs drawn from the WMT 2024 general shared task on machine translation. We then identify the primary triggers of verbosity, including safety, copyright concerns, and insufficient context in short input queries. Finally, we show that ignoring this behavior unfairly penalizes more verbose LLMs according to both automatic and human evaluations, highlighting the need to address this issue for more accurate future evaluations.","authors_text":"Colin Cherry, Eleftheria Briakou, Markus Freitag, Zhongtao Liu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-01T16:59:01Z","title":"On the Implications of Verbose LLM Outputs: A Case Study in Translation Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.00863","kind":"arxiv","version":1},"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":"sha256:6eede11272cee11c209aaa3440f60f87713ccab14a390aa4d59dc376162c8d70","target":"record","created_at":"2026-07-05T09:14:23Z","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":"d4c1722c9a3ed66b678aff471a40bb1251f325c7b87e3e5c0ae9e18a9a8e4e37","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-01T16:59:01Z","title_canon_sha256":"f13b168ff17ad1b960c32e9ca6e5c94dea3f8baa71e3f75fb9d44d5464bd4733"},"schema_version":"1.0","source":{"id":"2410.00863","kind":"arxiv","version":1}},"canonical_sha256":"c4cb24cf7806b7aa4396bd7c746a5d3bc6cf7a2b44f5b6b40fb85083f1db6a8b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c4cb24cf7806b7aa4396bd7c746a5d3bc6cf7a2b44f5b6b40fb85083f1db6a8b","first_computed_at":"2026-07-05T09:14:23.879818Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:14:23.879818Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"aw+9QPHn4w/Lbcy/9jjIe8IiAqxeR6UtrCCA4G10ouxA4OaySEnzFRqiktWYpTkriIdOQcJd4shKQGE7czHPCA==","signature_status":"signed_v1","signed_at":"2026-07-05T09:14:23.880323Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.00863","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6eede11272cee11c209aaa3440f60f87713ccab14a390aa4d59dc376162c8d70","sha256:8fc888225e820fb4da663ee4b1980e622bcfde6eb90d23f72d19904122949abc"],"state_sha256":"2d83ffe15f7b6fe7cbe30291592febfe772b86efd852a4670e6a3294b7928128"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aPsovkDzwdTleqN+cZY+wm4iExmjmts60FxsGUC4rjiRcE+sxznO/iud5tcmjtzmI3eSE3VeK9d0EdqQpoqmBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T13:27:06.777838Z","bundle_sha256":"f29e527202e3590cda7de9ed38c815b957db31af39c664a53bd6ca759655ceaf"}}