{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:AFBU3EZFBDF4FLVS3NLWZXFYNQ","short_pith_number":"pith:AFBU3EZF","schema_version":"1.0","canonical_sha256":"01434d932508cbc2aeb2db576cdcb86c38372400133e99ec29c243388ce2d812","source":{"kind":"arxiv","id":"2605.02443","version":2},"attestation_state":"computed","paper":{"title":"HalluScan: A Systematic Benchmark for Detecting and Mitigating Hallucinations in Instruction-Following LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Systematic benchmarking reveals NLI Verification as the most effective method for detecting hallucinations in LLMs at an AUROC of 0.88.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ahmed Cherif","submitted_at":"2026-05-04T10:43:27Z","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"},"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":true},"canonical_record":{"source":{"id":"2605.02443","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-04T10:43:27Z","cross_cats_sorted":[],"title_canon_sha256":"85d837d6891f48efbec948a0daf4484ec0c1b6bb65df91440679fdb993fb9319","abstract_canon_sha256":"a8d4aa271dcdba19a8e074e11c31c6c2314cc65b4663f6828f11fb536ca80ae9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-25T02:02:16.155138Z","signature_b64":"8QjU/Fg7RocVmKUB7k/La7mcYUrfButPjVIaSMFmjylwFvjiBoMoVe8Iw+2788zhhBjqqslD045wMyDAdMsKBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"01434d932508cbc2aeb2db576cdcb86c38372400133e99ec29c243388ce2d812","last_reissued_at":"2026-05-25T02:02:16.154597Z","signature_status":"signed_v1","first_computed_at":"2026-05-25T02:02:16.154597Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"HalluScan: A Systematic Benchmark for Detecting and Mitigating Hallucinations in Instruction-Following LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Systematic benchmarking reveals NLI Verification as the most effective method for detecting hallucinations in LLMs at an AUROC of 0.88.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ahmed Cherif","submitted_at":"2026-05-04T10:43:27Z","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"},"claims":{"count":4,"items":[{"kind":"strongest_claim","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.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","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.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","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.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Systematic benchmarking reveals NLI Verification as the most effective method for detecting hallucinations in LLMs at an AUROC of 0.88.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"f3a18cc1848e370773a4dc20497832d0bf7bf8819d4796e7cdab01b6c775c0d7"},"source":{"id":"2605.02443","kind":"arxiv","version":2},"verdict":{"id":"b7d1684b-537a-4e18-bf2b-4b834e1c50ee","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T18:25:03.167234Z","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.","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","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.","pith_extraction_headline":"Systematic benchmarking reveals NLI Verification as the most effective method for detecting hallucinations in LLMs at an AUROC of 0.88."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.02443/integrity.json","findings":[],"available":true,"detectors_run":[{"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","findings_count":0}],"snapshot_sha256":"b8c5b882ae4b8996fd4da6824b4d1c15004b86d4411de55f572f5a396c4a5592"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":3,"snapshot_sha256":"439e5e5d84d3042ab501687000b3249d1ced253ce7681d07e180d3c7a4a204bb"},"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":"2605.02443","created_at":"2026-05-25T02:02:16.154680+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.02443v2","created_at":"2026-05-25T02:02:16.154680+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.02443","created_at":"2026-05-25T02:02:16.154680+00:00"},{"alias_kind":"pith_short_12","alias_value":"AFBU3EZFBDF4","created_at":"2026-05-25T02:02:16.154680+00:00"},{"alias_kind":"pith_short_16","alias_value":"AFBU3EZFBDF4FLVS","created_at":"2026-05-25T02:02:16.154680+00:00"},{"alias_kind":"pith_short_8","alias_value":"AFBU3EZF","created_at":"2026-05-25T02:02:16.154680+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":3,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AFBU3EZFBDF4FLVS3NLWZXFYNQ","json":"https://pith.science/pith/AFBU3EZFBDF4FLVS3NLWZXFYNQ.json","graph_json":"https://pith.science/api/pith-number/AFBU3EZFBDF4FLVS3NLWZXFYNQ/graph.json","events_json":"https://pith.science/api/pith-number/AFBU3EZFBDF4FLVS3NLWZXFYNQ/events.json","paper":"https://pith.science/paper/AFBU3EZF"},"agent_actions":{"view_html":"https://pith.science/pith/AFBU3EZFBDF4FLVS3NLWZXFYNQ","download_json":"https://pith.science/pith/AFBU3EZFBDF4FLVS3NLWZXFYNQ.json","view_paper":"https://pith.science/paper/AFBU3EZF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.02443&json=true","fetch_graph":"https://pith.science/api/pith-number/AFBU3EZFBDF4FLVS3NLWZXFYNQ/graph.json","fetch_events":"https://pith.science/api/pith-number/AFBU3EZFBDF4FLVS3NLWZXFYNQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AFBU3EZFBDF4FLVS3NLWZXFYNQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AFBU3EZFBDF4FLVS3NLWZXFYNQ/action/storage_attestation","attest_author":"https://pith.science/pith/AFBU3EZFBDF4FLVS3NLWZXFYNQ/action/author_attestation","sign_citation":"https://pith.science/pith/AFBU3EZFBDF4FLVS3NLWZXFYNQ/action/citation_signature","submit_replication":"https://pith.science/pith/AFBU3EZFBDF4FLVS3NLWZXFYNQ/action/replication_record"}},"created_at":"2026-05-25T02:02:16.154680+00:00","updated_at":"2026-05-25T02:02:16.154680+00:00"}