{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:4N6V5UMLKPN2DUFB6LCP5ZX2VJ","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":"29580a1766070a52d191c145d93f95c7c4af2adf85053b3bd62663a6d10b13b4","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-18T07:10:03Z","title_canon_sha256":"39688f2f97b3068e2b8fcdb07cd327d8ef5d7754496b0fc7238c56935eb75109"},"schema_version":"1.0","source":{"id":"2505.12265","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.12265","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"arxiv_version","alias_value":"2505.12265v1","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12265","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"pith_short_12","alias_value":"4N6V5UMLKPN2","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"pith_short_16","alias_value":"4N6V5UMLKPN2DUFB","created_at":"2026-07-05T11:05:01Z"},{"alias_kind":"pith_short_8","alias_value":"4N6V5UML","created_at":"2026-07-05T11:05:01Z"}],"graph_snapshots":[{"event_id":"sha256:24fe6fa6659fbacf8b4d469c10b12642dff232978a594891c28bbab06632049e","target":"graph","created_at":"2026-07-05T11:05:01Z","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/2505.12265/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Hallucination, the generation of factually incorrect information, remains a significant challenge for large language models (LLMs), especially in open-domain long-form generation. Existing approaches for detecting hallucination in long-form tasks either focus on limited domains or rely heavily on external fact-checking tools, which may not always be available.\n  In this work, we systematically investigate reference-free hallucination detection in open-domain long-form responses. Our findings reveal that internal states (e.g., model's output probability and entropy) alone are insufficient for r","authors_text":"Aditya Tayade, Alexander Radovic, Arya Talebzadeh, Chengwei Qin, Eryk Helenowski, Han Fang, Hao Ma, Karthik Abinav Sankararaman, Nanshu Wang, Shafiq Joty, Sinong Wang, Tengyu Xu, Wenxuan Zhou","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-18T07:10:03Z","title":"Learning Auxiliary Tasks Improves Reference-Free Hallucination Detection in Open-Domain Long-Form Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12265","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":"sha256:6f8281db74351bd9afb443f86f455542549f489ffaca0b7dce4b22f753faf010","target":"record","created_at":"2026-07-05T11:05:01Z","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":"29580a1766070a52d191c145d93f95c7c4af2adf85053b3bd62663a6d10b13b4","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-18T07:10:03Z","title_canon_sha256":"39688f2f97b3068e2b8fcdb07cd327d8ef5d7754496b0fc7238c56935eb75109"},"schema_version":"1.0","source":{"id":"2505.12265","kind":"arxiv","version":1}},"canonical_sha256":"e37d5ed18b53dba1d0a1f2c4fee6faaa7d7fa6fa860375b39f4bd65a3ac81233","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e37d5ed18b53dba1d0a1f2c4fee6faaa7d7fa6fa860375b39f4bd65a3ac81233","first_computed_at":"2026-07-05T11:05:01.179512Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:05:01.179512Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"paa3lJ4Ybs7bIANtn3BjEDeek3BtSBqCJ05ts6wLm523JU77+odFvAY/Yjo0biQrJxqFWHY9Jc5msnjJw9suDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:05:01.179979Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.12265","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6f8281db74351bd9afb443f86f455542549f489ffaca0b7dce4b22f753faf010","sha256:24fe6fa6659fbacf8b4d469c10b12642dff232978a594891c28bbab06632049e"],"state_sha256":"cf8b2fe674fb563f46d33bbf1e901d5b76f288909b5eb5a34b87b5437af9ba1e"}