{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:ORWKCJAML2NAQJQ5CQRSP6G6RF","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":"61b0954995a3607aaa0a08aa62369064b504cf1b15bac486105b3c53bad0401c","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-09-04T07:33:44Z","title_canon_sha256":"fa6a6612ad0afe36823c218e1b51da348da3add0a9e05ce2f0bceb3596f82db5"},"schema_version":"1.0","source":{"id":"2509.03957","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.03957","created_at":"2026-07-05T12:04:50Z"},{"alias_kind":"arxiv_version","alias_value":"2509.03957v1","created_at":"2026-07-05T12:04:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.03957","created_at":"2026-07-05T12:04:50Z"},{"alias_kind":"pith_short_12","alias_value":"ORWKCJAML2NA","created_at":"2026-07-05T12:04:50Z"},{"alias_kind":"pith_short_16","alias_value":"ORWKCJAML2NAQJQ5","created_at":"2026-07-05T12:04:50Z"},{"alias_kind":"pith_short_8","alias_value":"ORWKCJAM","created_at":"2026-07-05T12:04:50Z"}],"graph_snapshots":[{"event_id":"sha256:39205b268ff7194716bf1d0f999633f16b53c45e53a516cce9ff2c152345df27","target":"graph","created_at":"2026-07-05T12:04:50Z","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/2509.03957/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The effectiveness of large language models (LLMs) to fact-check misinformation remains uncertain, despite their growing use. To this end, we present CANDY, a benchmark designed to systematically evaluate the capabilities and limitations of LLMs in fact-checking Chinese misinformation. Specifically, we curate a carefully annotated dataset of ~20k instances. Our analysis shows that current LLMs exhibit limitations in generating accurate fact-checking conclusions, even when enhanced with chain-of-thought reasoning and few-shot prompting. To understand these limitations, we develop a taxonomy to c","authors_text":"Chen Huang, Ruiling Guo, Tong Zhang, Xinwei Yang, Yong Hu","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-09-04T07:33:44Z","title":"CANDY: Benchmarking LLMs' Limitations and Assistive Potential in Chinese Misinformation Fact-Checking"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.03957","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":"sha256:920d01184578aa4f672c34031fd03b1b01309781fd02ecbd81798c36603b9bcd","target":"record","created_at":"2026-07-05T12:04:50Z","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":"61b0954995a3607aaa0a08aa62369064b504cf1b15bac486105b3c53bad0401c","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-09-04T07:33:44Z","title_canon_sha256":"fa6a6612ad0afe36823c218e1b51da348da3add0a9e05ce2f0bceb3596f82db5"},"schema_version":"1.0","source":{"id":"2509.03957","kind":"arxiv","version":1}},"canonical_sha256":"746ca1240c5e9a08261d142327f8de895ba1534cb15cac5162ddf84ec373dac0","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"746ca1240c5e9a08261d142327f8de895ba1534cb15cac5162ddf84ec373dac0","first_computed_at":"2026-07-05T12:04:50.418267Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T12:04:50.418267Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"TPa3PSOx3DEE8CmEzRH5KyBASmKhI8Vfs/jmVjSdwi36l2Vpc90hgREenghwAm687/7edxMNrfQbzDYm6GzpAg==","signature_status":"signed_v1","signed_at":"2026-07-05T12:04:50.418957Z","signed_message":"canonical_sha256_bytes"},"source_id":"2509.03957","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:920d01184578aa4f672c34031fd03b1b01309781fd02ecbd81798c36603b9bcd","sha256:39205b268ff7194716bf1d0f999633f16b53c45e53a516cce9ff2c152345df27"],"state_sha256":"96211e3be7519347b23ddf99179fb3d7e3d623c426a12eed23980a697cad073e"}