{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:LA3WM7EEQYMKBGVPFWO35IR4Z6","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":"cc51939423b0d40970bbf32c41c8bd632883e786c6707371080110bba1d4f56c","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-09-16T13:43:04Z","title_canon_sha256":"3eb5b2afcee424a29fe823b6b69cc5b596c58b8601ef122896d9ea4d2aa1d473"},"schema_version":"1.0","source":{"id":"2409.10280","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.10280","created_at":"2026-07-05T09:07:37Z"},{"alias_kind":"arxiv_version","alias_value":"2409.10280v1","created_at":"2026-07-05T09:07:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.10280","created_at":"2026-07-05T09:07:37Z"},{"alias_kind":"pith_short_12","alias_value":"LA3WM7EEQYMK","created_at":"2026-07-05T09:07:37Z"},{"alias_kind":"pith_short_16","alias_value":"LA3WM7EEQYMKBGVP","created_at":"2026-07-05T09:07:37Z"},{"alias_kind":"pith_short_8","alias_value":"LA3WM7EE","created_at":"2026-07-05T09:07:37Z"}],"graph_snapshots":[{"event_id":"sha256:9696c3e096d22ff83aba7dc87f9b6c7ff07d5ca8981f5b45c9f33f4c60ba2e78","target":"graph","created_at":"2026-07-05T09:07:37Z","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/2409.10280/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In recent years, the application of large language models (LLMs) to code-related tasks has gained significant attention. However, existing evaluation benchmarks often focus on limited scenarios, such as code generation or completion, which do not reflect the diverse challenges developers face in real-world contexts. To address this, we introduce ComplexCodeEval, a benchmark designed to assess LCMs in various development tasks, including code generation, completion, API recommendation, and test case generation. It includes 3,897 Java samples and 7,184 Python samples from high-star GitHub reposi","authors_text":"Chaozheng Wang, Chun Yong Chong, Cuiyun Gao, Jiachen Liu, Jia Feng, Shan Gao, Xin Xia","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-09-16T13:43:04Z","title":"ComplexCodeEval: A Benchmark for Evaluating Large Code Models on More Complex Code"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.10280","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":"sha256:5c53d41ed70f31ef799a420f6a222966b4105e215fee79b5bf4907e50dd522d6","target":"record","created_at":"2026-07-05T09:07:37Z","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":"cc51939423b0d40970bbf32c41c8bd632883e786c6707371080110bba1d4f56c","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SE","submitted_at":"2024-09-16T13:43:04Z","title_canon_sha256":"3eb5b2afcee424a29fe823b6b69cc5b596c58b8601ef122896d9ea4d2aa1d473"},"schema_version":"1.0","source":{"id":"2409.10280","kind":"arxiv","version":1}},"canonical_sha256":"5837667c848618a09aaf2d9dbea23ccf8add5e1605b65a960bf140589c34ab32","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5837667c848618a09aaf2d9dbea23ccf8add5e1605b65a960bf140589c34ab32","first_computed_at":"2026-07-05T09:07:37.797854Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:07:37.797854Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"RbfUJCfmzm7tOYS6abEKboVxiHQRY9z7EuFLOkCChpmN44vfOVFsjJ3C0GyjXWdYAhPSl0AiXYgUYHAhSgsNAw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:07:37.798344Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.10280","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5c53d41ed70f31ef799a420f6a222966b4105e215fee79b5bf4907e50dd522d6","sha256:9696c3e096d22ff83aba7dc87f9b6c7ff07d5ca8981f5b45c9f33f4c60ba2e78"],"state_sha256":"41faceab620421acec22bd10715fd7ad4f15eb86bc5f6748ac40e48905d17ede"}