{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:AFBU3EZFBDF4FLVS3NLWZXFYNQ","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":"a8d4aa271dcdba19a8e074e11c31c6c2314cc65b4663f6828f11fb536ca80ae9","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-04T10:43:27Z","title_canon_sha256":"85d837d6891f48efbec948a0daf4484ec0c1b6bb65df91440679fdb993fb9319"},"schema_version":"1.0","source":{"id":"2605.02443","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.02443","created_at":"2026-05-25T02:02:16Z"},{"alias_kind":"arxiv_version","alias_value":"2605.02443v2","created_at":"2026-05-25T02:02:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.02443","created_at":"2026-05-25T02:02:16Z"},{"alias_kind":"pith_short_12","alias_value":"AFBU3EZFBDF4","created_at":"2026-05-25T02:02:16Z"},{"alias_kind":"pith_short_16","alias_value":"AFBU3EZFBDF4FLVS","created_at":"2026-05-25T02:02:16Z"},{"alias_kind":"pith_short_8","alias_value":"AFBU3EZF","created_at":"2026-05-25T02:02:16Z"}],"graph_snapshots":[{"event_id":"sha256:77ded69ab2bc1b0af8a62cbd8117563e93e653ff010b5efdc1edef003fe2cafa","target":"graph","created_at":"2026-05-25T02:02:16Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"NLI Verification achieves the highest overall AUROC of 0.88, while RAV achieves the second-highest AUROC of 0.66. HalluScore achieves a Pearson correlation of r = 0.41 with human expert judgments."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the 72 configurations, chosen models, and three domains sufficiently represent the full range of hallucination behaviors in instruction-following LLMs and that human judgments provide a stable ground truth for validating the new metric."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"HalluScan benchmark tests hallucination detectors on LLMs, identifies NLI Verification as top performer with 0.88 AUROC, and introduces HalluScore (r=0.41 with humans) plus a routing method for 2x cost savings."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Systematic benchmarking reveals NLI Verification as the most effective method for detecting hallucinations in LLMs at an AUROC of 0.88."}],"snapshot_sha256":"f3a18cc1848e370773a4dc20497832d0bf7bf8819d4796e7cdab01b6c775c0d7"},"formal_canon":{"evidence_count":3,"snapshot_sha256":"439e5e5d84d3042ab501687000b3249d1ced253ce7681d07e180d3c7a4a204bb"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-20T15:40:33.758746Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-20T03:31:22.113923Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T16:22:47.877602Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2605.02443/integrity.json","findings":[],"snapshot_sha256":"b8c5b882ae4b8996fd4da6824b4d1c15004b86d4411de55f572f5a396c4a5592","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) have demonstrated remarkable capabilities across diverse natural language processing tasks, yet they remain susceptible to hallucinations -- generating content that is factually incorrect, unfaithful to provided context, or misaligned with user instructions. We present HalluScan, a comprehensive benchmark framework that systematically evaluates hallucination detection and mitigation across 72 configurations spanning 6 detection methods, 4 open-weight model families, and 3 diverse domains. We introduce three key contributions: (1) HalluScore, a novel composite metri","authors_text":"Ahmed Cherif","cross_cats":[],"headline":"Systematic benchmarking reveals NLI Verification as the most effective method for detecting hallucinations in LLMs at an AUROC of 0.88.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-04T10:43:27Z","title":"HalluScan: A Systematic Benchmark for Detecting and Mitigating Hallucinations in Instruction-Following LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.02443","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-08T18:25:03.167234Z","id":"b7d1684b-537a-4e18-bf2b-4b834e1c50ee","model_set":{"reader":"grok-4.3"},"one_line_summary":"HalluScan benchmark tests hallucination detectors on LLMs, identifies NLI Verification as top performer with 0.88 AUROC, and introduces HalluScore (r=0.41 with humans) plus a routing method for 2x cost savings.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Systematic benchmarking reveals NLI Verification as the most effective method for detecting hallucinations in LLMs at an AUROC of 0.88.","strongest_claim":"NLI Verification achieves the highest overall AUROC of 0.88, while RAV achieves the second-highest AUROC of 0.66. HalluScore achieves a Pearson correlation of r = 0.41 with human expert judgments.","weakest_assumption":"That the 72 configurations, chosen models, and three domains sufficiently represent the full range of hallucination behaviors in instruction-following LLMs and that human judgments provide a stable ground truth for validating the new metric."}},"verdict_id":"b7d1684b-537a-4e18-bf2b-4b834e1c50ee"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:a8d2fcacf4a9fcac7a384843a0d5c688db962fcc25ad54ac86f281e943affb22","target":"record","created_at":"2026-05-25T02:02:16Z","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":"a8d4aa271dcdba19a8e074e11c31c6c2314cc65b4663f6828f11fb536ca80ae9","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-04T10:43:27Z","title_canon_sha256":"85d837d6891f48efbec948a0daf4484ec0c1b6bb65df91440679fdb993fb9319"},"schema_version":"1.0","source":{"id":"2605.02443","kind":"arxiv","version":2}},"canonical_sha256":"01434d932508cbc2aeb2db576cdcb86c38372400133e99ec29c243388ce2d812","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"01434d932508cbc2aeb2db576cdcb86c38372400133e99ec29c243388ce2d812","first_computed_at":"2026-05-25T02:02:16.154597Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-25T02:02:16.154597Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8QjU/Fg7RocVmKUB7k/La7mcYUrfButPjVIaSMFmjylwFvjiBoMoVe8Iw+2788zhhBjqqslD045wMyDAdMsKBg==","signature_status":"signed_v1","signed_at":"2026-05-25T02:02:16.155138Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.02443","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a8d2fcacf4a9fcac7a384843a0d5c688db962fcc25ad54ac86f281e943affb22","sha256:77ded69ab2bc1b0af8a62cbd8117563e93e653ff010b5efdc1edef003fe2cafa"],"state_sha256":"377245a0ade9444e3d404ef84e3cf4f78fd403c8158062e593d0dcd33dadf33f"}