{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:7SOXCEXDI3PTKPZ6336K3J44V4","short_pith_number":"pith:7SOXCEXD","canonical_record":{"source":{"id":"2308.04945","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-08-09T13:22:37Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5f2284836787c11418a0c04fc7581acab12375396c0a7a0ff1c4c5665a013628","abstract_canon_sha256":"5211dc300d8f19b3db2b739fbfde8af95a30bd975b1b3f8d65fa29efc9dfe99c"},"schema_version":"1.0"},"canonical_sha256":"fc9d7112e346df353f3edefcada79caf01c244dfa03dfc58e4f2369402fc006f","source":{"kind":"arxiv","id":"2308.04945","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2308.04945","created_at":"2026-07-05T07:49:13Z"},{"alias_kind":"arxiv_version","alias_value":"2308.04945v2","created_at":"2026-07-05T07:49:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2308.04945","created_at":"2026-07-05T07:49:13Z"},{"alias_kind":"pith_short_12","alias_value":"7SOXCEXDI3PT","created_at":"2026-07-05T07:49:13Z"},{"alias_kind":"pith_short_16","alias_value":"7SOXCEXDI3PTKPZ6","created_at":"2026-07-05T07:49:13Z"},{"alias_kind":"pith_short_8","alias_value":"7SOXCEXD","created_at":"2026-07-05T07:49:13Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:7SOXCEXDI3PTKPZ6336K3J44V4","target":"record","payload":{"canonical_record":{"source":{"id":"2308.04945","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-08-09T13:22:37Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5f2284836787c11418a0c04fc7581acab12375396c0a7a0ff1c4c5665a013628","abstract_canon_sha256":"5211dc300d8f19b3db2b739fbfde8af95a30bd975b1b3f8d65fa29efc9dfe99c"},"schema_version":"1.0"},"canonical_sha256":"fc9d7112e346df353f3edefcada79caf01c244dfa03dfc58e4f2369402fc006f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:49:13.002477Z","signature_b64":"WiT7x8nWtSoTllMjyewo4tL+huhxMTDdtjtPfwhSGHzG52NGxYCckJhwTmxAyfuL95MyIXEZ1frSMW5og0eBCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fc9d7112e346df353f3edefcada79caf01c244dfa03dfc58e4f2369402fc006f","last_reissued_at":"2026-07-05T07:49:13.001996Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:49:13.001996Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2308.04945","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":"2026-07-05T07:49:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iZC4vEuCrWMy6Cw5FrqrdurqkZCT+Gr6GTkgSDrM37kG+CFWhbkhcbAUCqFhgLmj7pCdhDPeDuZK+plFqBHoDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T09:48:57.051603Z"},"content_sha256":"511824cc194c13ed8ce5a7c319d88d2606e2b5d8310a7973ff24e93f3357f042","schema_version":"1.0","event_id":"sha256:511824cc194c13ed8ce5a7c319d88d2606e2b5d8310a7973ff24e93f3357f042"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:7SOXCEXDI3PTKPZ6336K3J44V4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LLMeBench: A Flexible Framework for Accelerating LLMs Benchmarking","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Ahmed Abdelali, Ahmed Ali, Basel Mousi, Fahim Dalvi, Firoj Alam, Hamdy Mubarak, Majd Hawasly, Maram Hasanain, Nadir Durrani, Nizi Nazar, Sabri Boughorbel, Samir Abdaljalil, Shammur Absar Chowdhury","submitted_at":"2023-08-09T13:22:37Z","abstract_excerpt":"The recent development and success of Large Language Models (LLMs) necessitate an evaluation of their performance across diverse NLP tasks in different languages. Although several frameworks have been developed and made publicly available, their customization capabilities for specific tasks and datasets are often complex for different users. In this study, we introduce the LLMeBench framework, which can be seamlessly customized to evaluate LLMs for any NLP task, regardless of language. The framework features generic dataset loaders, several model providers, and pre-implements most standard eva"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2308.04945","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/2308.04945/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":"2026-07-05T07:49:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HpW+0xXqXKHsJqA77NbSuIWdyndbzLuC4PctaGt+7R7iBLWMTWZgW1v4mQGoA4bTg/2g5Jt6sc7aUZdYgmbYAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T09:48:57.052156Z"},"content_sha256":"7a3c04c58925bad735d5154c6ca947993e2aab9eb5ec2caeb4dd383aa14defa7","schema_version":"1.0","event_id":"sha256:7a3c04c58925bad735d5154c6ca947993e2aab9eb5ec2caeb4dd383aa14defa7"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/7SOXCEXDI3PTKPZ6336K3J44V4/bundle.json","state_url":"https://pith.science/pith/7SOXCEXDI3PTKPZ6336K3J44V4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/7SOXCEXDI3PTKPZ6336K3J44V4/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":"2026-08-18T09:48:57Z","links":{"resolver":"https://pith.science/pith/7SOXCEXDI3PTKPZ6336K3J44V4","bundle":"https://pith.science/pith/7SOXCEXDI3PTKPZ6336K3J44V4/bundle.json","state":"https://pith.science/pith/7SOXCEXDI3PTKPZ6336K3J44V4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/7SOXCEXDI3PTKPZ6336K3J44V4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:7SOXCEXDI3PTKPZ6336K3J44V4","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":"5211dc300d8f19b3db2b739fbfde8af95a30bd975b1b3f8d65fa29efc9dfe99c","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-08-09T13:22:37Z","title_canon_sha256":"5f2284836787c11418a0c04fc7581acab12375396c0a7a0ff1c4c5665a013628"},"schema_version":"1.0","source":{"id":"2308.04945","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2308.04945","created_at":"2026-07-05T07:49:13Z"},{"alias_kind":"arxiv_version","alias_value":"2308.04945v2","created_at":"2026-07-05T07:49:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2308.04945","created_at":"2026-07-05T07:49:13Z"},{"alias_kind":"pith_short_12","alias_value":"7SOXCEXDI3PT","created_at":"2026-07-05T07:49:13Z"},{"alias_kind":"pith_short_16","alias_value":"7SOXCEXDI3PTKPZ6","created_at":"2026-07-05T07:49:13Z"},{"alias_kind":"pith_short_8","alias_value":"7SOXCEXD","created_at":"2026-07-05T07:49:13Z"}],"graph_snapshots":[{"event_id":"sha256:7a3c04c58925bad735d5154c6ca947993e2aab9eb5ec2caeb4dd383aa14defa7","target":"graph","created_at":"2026-07-05T07:49:13Z","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/2308.04945/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The recent development and success of Large Language Models (LLMs) necessitate an evaluation of their performance across diverse NLP tasks in different languages. Although several frameworks have been developed and made publicly available, their customization capabilities for specific tasks and datasets are often complex for different users. In this study, we introduce the LLMeBench framework, which can be seamlessly customized to evaluate LLMs for any NLP task, regardless of language. The framework features generic dataset loaders, several model providers, and pre-implements most standard eva","authors_text":"Ahmed Abdelali, Ahmed Ali, Basel Mousi, Fahim Dalvi, Firoj Alam, Hamdy Mubarak, Majd Hawasly, Maram Hasanain, Nadir Durrani, Nizi Nazar, Sabri Boughorbel, Samir Abdaljalil, Shammur Absar Chowdhury","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-08-09T13:22:37Z","title":"LLMeBench: A Flexible Framework for Accelerating LLMs Benchmarking"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2308.04945","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":"sha256:511824cc194c13ed8ce5a7c319d88d2606e2b5d8310a7973ff24e93f3357f042","target":"record","created_at":"2026-07-05T07:49:13Z","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":"5211dc300d8f19b3db2b739fbfde8af95a30bd975b1b3f8d65fa29efc9dfe99c","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-08-09T13:22:37Z","title_canon_sha256":"5f2284836787c11418a0c04fc7581acab12375396c0a7a0ff1c4c5665a013628"},"schema_version":"1.0","source":{"id":"2308.04945","kind":"arxiv","version":2}},"canonical_sha256":"fc9d7112e346df353f3edefcada79caf01c244dfa03dfc58e4f2369402fc006f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"fc9d7112e346df353f3edefcada79caf01c244dfa03dfc58e4f2369402fc006f","first_computed_at":"2026-07-05T07:49:13.001996Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:49:13.001996Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"WiT7x8nWtSoTllMjyewo4tL+huhxMTDdtjtPfwhSGHzG52NGxYCckJhwTmxAyfuL95MyIXEZ1frSMW5og0eBCg==","signature_status":"signed_v1","signed_at":"2026-07-05T07:49:13.002477Z","signed_message":"canonical_sha256_bytes"},"source_id":"2308.04945","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:511824cc194c13ed8ce5a7c319d88d2606e2b5d8310a7973ff24e93f3357f042","sha256:7a3c04c58925bad735d5154c6ca947993e2aab9eb5ec2caeb4dd383aa14defa7"],"state_sha256":"60f857e654fa0c1c0489d6af7e420dc8fa624e43287ff9eb0da6861a1abdf1c8"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"vkxuGbi05VBtPzdXAn1dGbydF0X21TLGx6ZAh4fp58CNlRp6R49Z9nReEuPr5VxMpP7cvC2PccoizHGaDI4NBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-18T09:48:57.057426Z","bundle_sha256":"70477a6f6d657aa496325e8d2050893c7870220f0b5fcfceeab8346ff334e867"}}