{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:STRIJ3LLUEI5WGJI3JXRL36XD7","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":"853b3a352737e30905fe1dd4b4c332862f81d19f169d1162c01beeaf2f8c6848","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-17T11:41:53Z","title_canon_sha256":"9ea33b2b08008b21de098db752427dc13526583b79b1d798f034287cb8f21568"},"schema_version":"1.0","source":{"id":"2501.10132","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.10132","created_at":"2026-07-05T10:02:17Z"},{"alias_kind":"arxiv_version","alias_value":"2501.10132v1","created_at":"2026-07-05T10:02:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.10132","created_at":"2026-07-05T10:02:17Z"},{"alias_kind":"pith_short_12","alias_value":"STRIJ3LLUEI5","created_at":"2026-07-05T10:02:17Z"},{"alias_kind":"pith_short_16","alias_value":"STRIJ3LLUEI5WGJI","created_at":"2026-07-05T10:02:17Z"},{"alias_kind":"pith_short_8","alias_value":"STRIJ3LL","created_at":"2026-07-05T10:02:17Z"}],"graph_snapshots":[{"event_id":"sha256:7e04467a6b69370f95502f1f3ce7c47186631d421cd55cfb43ee0bcecce9999b","target":"graph","created_at":"2026-07-05T10:02:17Z","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/2501.10132/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Enhancing large language models (LLMs) with real-time APIs can help generate more accurate and up-to-date responses. However, evaluating the function calling abilities of LLMs in real-world scenarios remains under-explored due to the complexity of data collection and evaluation. In this work, we introduce ComplexFuncBench, a benchmark for complex function calling across five real-world scenarios. Compared to existing benchmarks, ComplexFuncBench encompasses multi-step and constrained function calling, which requires long-parameter filing, parameter value reasoning, and 128k long context. Addit","authors_text":"Haiyi Hu, Jie Tang, Lucen Zhong, Xiaohan Zhang, Zhengxiao Du","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-17T11:41:53Z","title":"ComplexFuncBench: Exploring Multi-Step and Constrained Function Calling under Long-Context Scenario"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.10132","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":"sha256:43331025d7ec4a7566e4e00d38bd319408e2bb2441c1e457fe008239e1ecd0c2","target":"record","created_at":"2026-07-05T10:02:17Z","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":"853b3a352737e30905fe1dd4b4c332862f81d19f169d1162c01beeaf2f8c6848","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-17T11:41:53Z","title_canon_sha256":"9ea33b2b08008b21de098db752427dc13526583b79b1d798f034287cb8f21568"},"schema_version":"1.0","source":{"id":"2501.10132","kind":"arxiv","version":1}},"canonical_sha256":"94e284ed6ba111db1928da6f15efd71fccc979bedbde2a1e3e0db1fef77a880f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"94e284ed6ba111db1928da6f15efd71fccc979bedbde2a1e3e0db1fef77a880f","first_computed_at":"2026-07-05T10:02:17.261290Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:02:17.261290Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"TuvTCpn5txD4hr7RN3oC5xc6pHS4345rLf/5n8PiTtglTdfrrjsM50SNbB1E9x8kJCInpXMpxEfJXcMbMH7JDA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:02:17.261726Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.10132","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:43331025d7ec4a7566e4e00d38bd319408e2bb2441c1e457fe008239e1ecd0c2","sha256:7e04467a6b69370f95502f1f3ce7c47186631d421cd55cfb43ee0bcecce9999b"],"state_sha256":"ad119b329a4fc37631ebec2c25a061a5dddea480cb41a7731178f1b70118f576"}