{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:MYWQKVUNKOZ2OPECCDBVZ7TSAK","short_pith_number":"pith:MYWQKVUN","canonical_record":{"source":{"id":"2308.04813","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-08-09T09:11:31Z","cross_cats_sorted":[],"title_canon_sha256":"f57a7f418c95162cf835261d153b6c650a29f5b7622f6717ebee6b0ecd4aadfb","abstract_canon_sha256":"65ba989615ec7275fe2b71713bd81e2e263a805b19b8a51b43866c22307c1651"},"schema_version":"1.0"},"canonical_sha256":"662d05568d53b3a73c8210c35cfe7202bb6cb2a8eb3f654668d980d91157062a","source":{"kind":"arxiv","id":"2308.04813","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2308.04813","created_at":"2026-07-05T07:01:04Z"},{"alias_kind":"arxiv_version","alias_value":"2308.04813v2","created_at":"2026-07-05T07:01:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2308.04813","created_at":"2026-07-05T07:01:04Z"},{"alias_kind":"pith_short_12","alias_value":"MYWQKVUNKOZ2","created_at":"2026-07-05T07:01:04Z"},{"alias_kind":"pith_short_16","alias_value":"MYWQKVUNKOZ2OPEC","created_at":"2026-07-05T07:01:04Z"},{"alias_kind":"pith_short_8","alias_value":"MYWQKVUN","created_at":"2026-07-05T07:01:04Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:MYWQKVUNKOZ2OPECCDBVZ7TSAK","target":"record","payload":{"canonical_record":{"source":{"id":"2308.04813","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-08-09T09:11:31Z","cross_cats_sorted":[],"title_canon_sha256":"f57a7f418c95162cf835261d153b6c650a29f5b7622f6717ebee6b0ecd4aadfb","abstract_canon_sha256":"65ba989615ec7275fe2b71713bd81e2e263a805b19b8a51b43866c22307c1651"},"schema_version":"1.0"},"canonical_sha256":"662d05568d53b3a73c8210c35cfe7202bb6cb2a8eb3f654668d980d91157062a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:01:04.389554Z","signature_b64":"Reu6YPyQDJEqH0DBw0F0ZZoA8A9hMjtIq8Iew0/HTwtkFq0CmSaScxgag5Qb5WrRwRasUZXXSCNL6gcC9z7DDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"662d05568d53b3a73c8210c35cfe7202bb6cb2a8eb3f654668d980d91157062a","last_reissued_at":"2026-07-05T07:01:04.389066Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:01:04.389066Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2308.04813","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":"2026-07-05T07:01:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HIVWbzYWvTTxATDJONa5jkqOSwI26+XYWgvQh/fIZCxC3pmoT1e+Nir81HRdnbTu4+UJ9whpYcxgGs7vVCWMCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T22:31:04.372344Z"},"content_sha256":"aa98100c454c17be979dc36116a7c9ddf09ec493da442e0e7935b2b40d32d78e","schema_version":"1.0","event_id":"sha256:aa98100c454c17be979dc36116a7c9ddf09ec493da442e0e7935b2b40d32d78e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:MYWQKVUNKOZ2OPECCDBVZ7TSAK","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"CLEVA: Chinese Language Models EVAluation Platform","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dahua Lin, Duo Zheng, Jianqiao Zhao, Liwei Wang, Michael R. Lyu, Shijia Huang, Xiaohui Su, Yanyang Li, Yongfeng Huang, Zhi Chen, Zi-Yuan Hu","submitted_at":"2023-08-09T09:11:31Z","abstract_excerpt":"With the continuous emergence of Chinese Large Language Models (LLMs), how to evaluate a model's capabilities has become an increasingly significant issue. The absence of a comprehensive Chinese benchmark that thoroughly assesses a model's performance, the unstandardized and incomparable prompting procedure, and the prevalent risk of contamination pose major challenges in the current evaluation of Chinese LLMs. We present CLEVA, a user-friendly platform crafted to holistically evaluate Chinese LLMs. Our platform employs a standardized workflow to assess LLMs' performance across various dimensi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2308.04813","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/2308.04813/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":"2026-07-05T07:01:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"L1nH95gyYFHMQOD582Z0yhBq70DxgHJqRcAVczoDIQUfmU7zqSoeaIMPMxYAxIJ8qVmgTC/LKHDU7cxviIiVAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T22:31:04.372919Z"},"content_sha256":"0776701af49b03aaee6df68988f729e4c5042c6a27b4e5f1adb45e2c6753fcf9","schema_version":"1.0","event_id":"sha256:0776701af49b03aaee6df68988f729e4c5042c6a27b4e5f1adb45e2c6753fcf9"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MYWQKVUNKOZ2OPECCDBVZ7TSAK/bundle.json","state_url":"https://pith.science/pith/MYWQKVUNKOZ2OPECCDBVZ7TSAK/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MYWQKVUNKOZ2OPECCDBVZ7TSAK/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":"2026-08-03T22:31:04Z","links":{"resolver":"https://pith.science/pith/MYWQKVUNKOZ2OPECCDBVZ7TSAK","bundle":"https://pith.science/pith/MYWQKVUNKOZ2OPECCDBVZ7TSAK/bundle.json","state":"https://pith.science/pith/MYWQKVUNKOZ2OPECCDBVZ7TSAK/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MYWQKVUNKOZ2OPECCDBVZ7TSAK/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:MYWQKVUNKOZ2OPECCDBVZ7TSAK","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":"65ba989615ec7275fe2b71713bd81e2e263a805b19b8a51b43866c22307c1651","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-08-09T09:11:31Z","title_canon_sha256":"f57a7f418c95162cf835261d153b6c650a29f5b7622f6717ebee6b0ecd4aadfb"},"schema_version":"1.0","source":{"id":"2308.04813","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2308.04813","created_at":"2026-07-05T07:01:04Z"},{"alias_kind":"arxiv_version","alias_value":"2308.04813v2","created_at":"2026-07-05T07:01:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2308.04813","created_at":"2026-07-05T07:01:04Z"},{"alias_kind":"pith_short_12","alias_value":"MYWQKVUNKOZ2","created_at":"2026-07-05T07:01:04Z"},{"alias_kind":"pith_short_16","alias_value":"MYWQKVUNKOZ2OPEC","created_at":"2026-07-05T07:01:04Z"},{"alias_kind":"pith_short_8","alias_value":"MYWQKVUN","created_at":"2026-07-05T07:01:04Z"}],"graph_snapshots":[{"event_id":"sha256:0776701af49b03aaee6df68988f729e4c5042c6a27b4e5f1adb45e2c6753fcf9","target":"graph","created_at":"2026-07-05T07:01:04Z","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/2308.04813/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"With the continuous emergence of Chinese Large Language Models (LLMs), how to evaluate a model's capabilities has become an increasingly significant issue. The absence of a comprehensive Chinese benchmark that thoroughly assesses a model's performance, the unstandardized and incomparable prompting procedure, and the prevalent risk of contamination pose major challenges in the current evaluation of Chinese LLMs. We present CLEVA, a user-friendly platform crafted to holistically evaluate Chinese LLMs. Our platform employs a standardized workflow to assess LLMs' performance across various dimensi","authors_text":"Dahua Lin, Duo Zheng, Jianqiao Zhao, Liwei Wang, Michael R. Lyu, Shijia Huang, Xiaohui Su, Yanyang Li, Yongfeng Huang, Zhi Chen, Zi-Yuan Hu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-08-09T09:11:31Z","title":"CLEVA: Chinese Language Models EVAluation Platform"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2308.04813","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:aa98100c454c17be979dc36116a7c9ddf09ec493da442e0e7935b2b40d32d78e","target":"record","created_at":"2026-07-05T07:01:04Z","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":"65ba989615ec7275fe2b71713bd81e2e263a805b19b8a51b43866c22307c1651","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-08-09T09:11:31Z","title_canon_sha256":"f57a7f418c95162cf835261d153b6c650a29f5b7622f6717ebee6b0ecd4aadfb"},"schema_version":"1.0","source":{"id":"2308.04813","kind":"arxiv","version":2}},"canonical_sha256":"662d05568d53b3a73c8210c35cfe7202bb6cb2a8eb3f654668d980d91157062a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"662d05568d53b3a73c8210c35cfe7202bb6cb2a8eb3f654668d980d91157062a","first_computed_at":"2026-07-05T07:01:04.389066Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:01:04.389066Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Reu6YPyQDJEqH0DBw0F0ZZoA8A9hMjtIq8Iew0/HTwtkFq0CmSaScxgag5Qb5WrRwRasUZXXSCNL6gcC9z7DDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T07:01:04.389554Z","signed_message":"canonical_sha256_bytes"},"source_id":"2308.04813","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:aa98100c454c17be979dc36116a7c9ddf09ec493da442e0e7935b2b40d32d78e","sha256:0776701af49b03aaee6df68988f729e4c5042c6a27b4e5f1adb45e2c6753fcf9"],"state_sha256":"da4554eee55b83a1d37c3520159959b585842aa5a1b492a6122cd637cdd82507"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"euTCPnY3oVriyK/rvMIsilTtmSXqtNjskKYCE59/4CRRKp8tX08r0UpUpKVWXW8R1GnrNIj4bziABKRcnvcjCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-03T22:31:04.376950Z","bundle_sha256":"fcc442d526da7e915cda8c3fb37ee9d9e770a1fe90c6719f2deb24ee35ae8e7d"}}