{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:KVG2R2BRDODI6SGZOWNY6TX3JD","short_pith_number":"pith:KVG2R2BR","schema_version":"1.0","canonical_sha256":"554da8e8311b868f48d9759b8f4efb48e5c1912c23c3e717a378e866f2b7e05f","source":{"kind":"arxiv","id":"2608.07838","version":1},"attestation_state":"computed","paper":{"title":"Counterfactual Benchmarking and Training for Factuality Consistency and Order-Robust Grounded Reasoning in LLMs over Heterogeneous Knowledge","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Haofen Wang, Lianghua He, Shibo Chu, Tiehua Zhang, Yuze Liu, Zhijun Ding, Zhishu Shen","submitted_at":"2026-08-08T00:46:32Z","abstract_excerpt":"Large language models (LLMs) have increasingly supported response generation grounded in user-provided knowledge spanning heterogeneous structures. However, existing benchmarks provide limited assessment of whether LLMs can faithfully perform multi-hop reasoning chains across such knowledge contexts while remaining robust to variations in their input order. We introduce TKFQA, a factuality consistency benchmark comprising 10,130 question-answering (QA) pairs grounded in tables, texts, and knowledge graphs (KGs). Each example is constructed from an explicit counterfactual reasoning chain, enabl"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2608.07838","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-08-08T00:46:32Z","cross_cats_sorted":[],"title_canon_sha256":"70a03f5643aaa221c86753504ed18b20eae07e1f2d14a403b0a26f41b98806ac","abstract_canon_sha256":"eb8bd3aa57c15d95fa8bfdfc3b67a3f0982b2c326386d174ef4d794b64173d48"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T01:18:38.958213Z","signature_b64":"ICgiZkX8esGotRsmQYe1dGWVW/JqIU2jNDolp1FwOI3SKwu4cFO7KKnE8uHbxVwqaUc0EhjOzNY7HmvFynlTDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"554da8e8311b868f48d9759b8f4efb48e5c1912c23c3e717a378e866f2b7e05f","last_reissued_at":"2026-08-11T01:18:38.955677Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T01:18:38.955677Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Counterfactual Benchmarking and Training for Factuality Consistency and Order-Robust Grounded Reasoning in LLMs over Heterogeneous Knowledge","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Haofen Wang, Lianghua He, Shibo Chu, Tiehua Zhang, Yuze Liu, Zhijun Ding, Zhishu Shen","submitted_at":"2026-08-08T00:46:32Z","abstract_excerpt":"Large language models (LLMs) have increasingly supported response generation grounded in user-provided knowledge spanning heterogeneous structures. However, existing benchmarks provide limited assessment of whether LLMs can faithfully perform multi-hop reasoning chains across such knowledge contexts while remaining robust to variations in their input order. We introduce TKFQA, a factuality consistency benchmark comprising 10,130 question-answering (QA) pairs grounded in tables, texts, and knowledge graphs (KGs). Each example is constructed from an explicit counterfactual reasoning chain, enabl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.07838","kind":"arxiv","version":1},"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/2608.07838/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2608.07838","created_at":"2026-08-11T01:18:38.956427+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.07838v1","created_at":"2026-08-11T01:18:38.956427+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.07838","created_at":"2026-08-11T01:18:38.956427+00:00"},{"alias_kind":"pith_short_12","alias_value":"KVG2R2BRDODI","created_at":"2026-08-11T01:18:38.956427+00:00"},{"alias_kind":"pith_short_16","alias_value":"KVG2R2BRDODI6SGZ","created_at":"2026-08-11T01:18:38.956427+00:00"},{"alias_kind":"pith_short_8","alias_value":"KVG2R2BR","created_at":"2026-08-11T01:18:38.956427+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KVG2R2BRDODI6SGZOWNY6TX3JD","json":"https://pith.science/pith/KVG2R2BRDODI6SGZOWNY6TX3JD.json","graph_json":"https://pith.science/api/pith-number/KVG2R2BRDODI6SGZOWNY6TX3JD/graph.json","events_json":"https://pith.science/api/pith-number/KVG2R2BRDODI6SGZOWNY6TX3JD/events.json","paper":"https://pith.science/paper/KVG2R2BR"},"agent_actions":{"view_html":"https://pith.science/pith/KVG2R2BRDODI6SGZOWNY6TX3JD","download_json":"https://pith.science/pith/KVG2R2BRDODI6SGZOWNY6TX3JD.json","view_paper":"https://pith.science/paper/KVG2R2BR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.07838&json=true","fetch_graph":"https://pith.science/api/pith-number/KVG2R2BRDODI6SGZOWNY6TX3JD/graph.json","fetch_events":"https://pith.science/api/pith-number/KVG2R2BRDODI6SGZOWNY6TX3JD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KVG2R2BRDODI6SGZOWNY6TX3JD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KVG2R2BRDODI6SGZOWNY6TX3JD/action/storage_attestation","attest_author":"https://pith.science/pith/KVG2R2BRDODI6SGZOWNY6TX3JD/action/author_attestation","sign_citation":"https://pith.science/pith/KVG2R2BRDODI6SGZOWNY6TX3JD/action/citation_signature","submit_replication":"https://pith.science/pith/KVG2R2BRDODI6SGZOWNY6TX3JD/action/replication_record"}},"created_at":"2026-08-11T01:18:38.956427+00:00","updated_at":"2026-08-11T01:18:38.956427+00:00"}