{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:JKBJXSVC7VBWVW7GQAPK7WUN7H","short_pith_number":"pith:JKBJXSVC","canonical_record":{"source":{"id":"2504.11829","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-16T07:38:19Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ef592b4c2824f3236174037d91f319f5e09d5aadf1c3a91954fbe52e2d6fffb1","abstract_canon_sha256":"f288d78f73beb3d015a6768962156258a950cfa3f0df7a40c470b401a2437e85"},"schema_version":"1.0"},"canonical_sha256":"4a829bcaa2fd436adbe6801eafda8df9df893ffcc8e0f30faac8fd56e134713c","source":{"kind":"arxiv","id":"2504.11829","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.11829","created_at":"2026-07-05T12:09:44Z"},{"alias_kind":"arxiv_version","alias_value":"2504.11829v4","created_at":"2026-07-05T12:09:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.11829","created_at":"2026-07-05T12:09:44Z"},{"alias_kind":"pith_short_12","alias_value":"JKBJXSVC7VBW","created_at":"2026-07-05T12:09:44Z"},{"alias_kind":"pith_short_16","alias_value":"JKBJXSVC7VBWVW7G","created_at":"2026-07-05T12:09:44Z"},{"alias_kind":"pith_short_8","alias_value":"JKBJXSVC","created_at":"2026-07-05T12:09:44Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:JKBJXSVC7VBWVW7GQAPK7WUN7H","target":"record","payload":{"canonical_record":{"source":{"id":"2504.11829","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-16T07:38:19Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ef592b4c2824f3236174037d91f319f5e09d5aadf1c3a91954fbe52e2d6fffb1","abstract_canon_sha256":"f288d78f73beb3d015a6768962156258a950cfa3f0df7a40c470b401a2437e85"},"schema_version":"1.0"},"canonical_sha256":"4a829bcaa2fd436adbe6801eafda8df9df893ffcc8e0f30faac8fd56e134713c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:09:44.114340Z","signature_b64":"T0WQHAWlyY3PswEn+8jVHO9T2MTkS3mbsPRFkfjb0XYwVosg80b+NzSAfPg+ML0Y3rPzjNHOk9xG7NueV0+6CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4a829bcaa2fd436adbe6801eafda8df9df893ffcc8e0f30faac8fd56e134713c","last_reissued_at":"2026-07-05T12:09:44.113810Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:09:44.113810Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2504.11829","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T12:09:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Q5oWZqobBIp/jLci+GSG+PCyAJx5lHBfV1Ts6t/xwE5exuLq3teYUr9vq/JPQ6020HByv6+LAggmef1nG/0CBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T09:39:43.603588Z"},"content_sha256":"c4ed6473545599a409e911e988bcd5646d340a93322fa93f287f91478a5a18c5","schema_version":"1.0","event_id":"sha256:c4ed6473545599a409e911e988bcd5646d340a93322fa93f287f91478a5a18c5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:JKBJXSVC7VBWVW7GQAPK7WUN7H","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"D\\'ej\\`a Vu: Multilingual LLM Evaluation through the Lens of Machine Translation Evaluation","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Eleftheria Briakou, Julia Kreutzer, Kocmi Tom, Marzieh Fadaee, Sweta Agrawal","submitted_at":"2025-04-16T07:38:19Z","abstract_excerpt":"Generation capabilities and language coverage of multilingual large language models (mLLMs) are advancing rapidly. However, evaluation practices for generative abilities of mLLMs are still lacking comprehensiveness, scientific rigor, and consistent adoption across research labs, which undermines their potential to meaningfully guide mLLM development. We draw parallels with machine translation (MT) evaluation, a field that faced similar challenges and has, over decades, developed transparent reporting standards and reliable evaluations for multilingual generative models. Through targeted experi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.11829","kind":"arxiv","version":4},"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/2504.11829/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":"2026-07-05T12:09:44Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"sRMak0c0IwdLQE/4ibldoi/v0UNbCXgMFWEG6W9GpsuLnAR3x7btp70sle54OsKeWKsMASMNhB+wKdC1oNiTDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T09:39:43.604202Z"},"content_sha256":"06c37e8e9231236cf02b03bda65eb9015ff4ca381c9a4dcfe97e67021ae3c8f9","schema_version":"1.0","event_id":"sha256:06c37e8e9231236cf02b03bda65eb9015ff4ca381c9a4dcfe97e67021ae3c8f9"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/JKBJXSVC7VBWVW7GQAPK7WUN7H/bundle.json","state_url":"https://pith.science/pith/JKBJXSVC7VBWVW7GQAPK7WUN7H/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/JKBJXSVC7VBWVW7GQAPK7WUN7H/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":"2026-08-08T09:39:43Z","links":{"resolver":"https://pith.science/pith/JKBJXSVC7VBWVW7GQAPK7WUN7H","bundle":"https://pith.science/pith/JKBJXSVC7VBWVW7GQAPK7WUN7H/bundle.json","state":"https://pith.science/pith/JKBJXSVC7VBWVW7GQAPK7WUN7H/state.json","well_known_bundle":"https://pith.science/.well-known/pith/JKBJXSVC7VBWVW7GQAPK7WUN7H/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:JKBJXSVC7VBWVW7GQAPK7WUN7H","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":"f288d78f73beb3d015a6768962156258a950cfa3f0df7a40c470b401a2437e85","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-16T07:38:19Z","title_canon_sha256":"ef592b4c2824f3236174037d91f319f5e09d5aadf1c3a91954fbe52e2d6fffb1"},"schema_version":"1.0","source":{"id":"2504.11829","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.11829","created_at":"2026-07-05T12:09:44Z"},{"alias_kind":"arxiv_version","alias_value":"2504.11829v4","created_at":"2026-07-05T12:09:44Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.11829","created_at":"2026-07-05T12:09:44Z"},{"alias_kind":"pith_short_12","alias_value":"JKBJXSVC7VBW","created_at":"2026-07-05T12:09:44Z"},{"alias_kind":"pith_short_16","alias_value":"JKBJXSVC7VBWVW7G","created_at":"2026-07-05T12:09:44Z"},{"alias_kind":"pith_short_8","alias_value":"JKBJXSVC","created_at":"2026-07-05T12:09:44Z"}],"graph_snapshots":[{"event_id":"sha256:06c37e8e9231236cf02b03bda65eb9015ff4ca381c9a4dcfe97e67021ae3c8f9","target":"graph","created_at":"2026-07-05T12:09:44Z","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/2504.11829/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Generation capabilities and language coverage of multilingual large language models (mLLMs) are advancing rapidly. However, evaluation practices for generative abilities of mLLMs are still lacking comprehensiveness, scientific rigor, and consistent adoption across research labs, which undermines their potential to meaningfully guide mLLM development. We draw parallels with machine translation (MT) evaluation, a field that faced similar challenges and has, over decades, developed transparent reporting standards and reliable evaluations for multilingual generative models. Through targeted experi","authors_text":"Eleftheria Briakou, Julia Kreutzer, Kocmi Tom, Marzieh Fadaee, Sweta Agrawal","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-16T07:38:19Z","title":"D\\'ej\\`a Vu: Multilingual LLM Evaluation through the Lens of Machine Translation Evaluation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.11829","kind":"arxiv","version":4},"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:c4ed6473545599a409e911e988bcd5646d340a93322fa93f287f91478a5a18c5","target":"record","created_at":"2026-07-05T12:09:44Z","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":"f288d78f73beb3d015a6768962156258a950cfa3f0df7a40c470b401a2437e85","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-16T07:38:19Z","title_canon_sha256":"ef592b4c2824f3236174037d91f319f5e09d5aadf1c3a91954fbe52e2d6fffb1"},"schema_version":"1.0","source":{"id":"2504.11829","kind":"arxiv","version":4}},"canonical_sha256":"4a829bcaa2fd436adbe6801eafda8df9df893ffcc8e0f30faac8fd56e134713c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4a829bcaa2fd436adbe6801eafda8df9df893ffcc8e0f30faac8fd56e134713c","first_computed_at":"2026-07-05T12:09:44.113810Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T12:09:44.113810Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"T0WQHAWlyY3PswEn+8jVHO9T2MTkS3mbsPRFkfjb0XYwVosg80b+NzSAfPg+ML0Y3rPzjNHOk9xG7NueV0+6CA==","signature_status":"signed_v1","signed_at":"2026-07-05T12:09:44.114340Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.11829","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c4ed6473545599a409e911e988bcd5646d340a93322fa93f287f91478a5a18c5","sha256:06c37e8e9231236cf02b03bda65eb9015ff4ca381c9a4dcfe97e67021ae3c8f9"],"state_sha256":"3eb08a686084a41e9bb3cafdec691ffd94c9a2f4b3a1846d7c24271fbc979954"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RxRYwHXaVvU/6ULB6n/XWc0bpOCaHMQkILyKXZjaVtfkTWJTdgy92ZPDzzV+7dailSjZHfLavDKpxXOeU1kxAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T09:39:43.609012Z","bundle_sha256":"1d802ae40b42afce5ea104b9cdf044095a9c7bcc145376a43db3f9ac50a852c0"}}