{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:RDBHOTOZCW24G7ZRPYMM5LL6AS","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":"a91ab13e23b13092085525462d6d046559af57a55f7585d5e8a29dd258de43d7","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CY","submitted_at":"2025-05-29T17:13:26Z","title_canon_sha256":"4859750b1e6dff2bec7874b17228fc3c3270e9d26b2e0ebfe3458d06b8e5de2d"},"schema_version":"1.0","source":{"id":"2506.00072","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.00072","created_at":"2026-07-05T11:13:28Z"},{"alias_kind":"arxiv_version","alias_value":"2506.00072v1","created_at":"2026-07-05T11:13:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.00072","created_at":"2026-07-05T11:13:28Z"},{"alias_kind":"pith_short_12","alias_value":"RDBHOTOZCW24","created_at":"2026-07-05T11:13:28Z"},{"alias_kind":"pith_short_16","alias_value":"RDBHOTOZCW24G7ZR","created_at":"2026-07-05T11:13:28Z"},{"alias_kind":"pith_short_8","alias_value":"RDBHOTOZ","created_at":"2026-07-05T11:13:28Z"}],"graph_snapshots":[{"event_id":"sha256:f29b14e65222b2e4f8f5abb9b04f2fd0673a37317e94970177215ab9dfb3bdbd","target":"graph","created_at":"2026-07-05T11:13:28Z","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/2506.00072/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper investigates how prompt engineering techniques impact both accuracy and confidence elicitation in Large Language Models (LLMs) applied to medical contexts. Using a stratified dataset of Persian board exam questions across multiple specialties, we evaluated five LLMs - GPT-4o, o3-mini, Llama-3.3-70b, Llama-3.1-8b, and DeepSeek-v3 - across 156 configurations. These configurations varied in temperature settings (0.3, 0.7, 1.0), prompt styles (Chain-of-Thought, Few-Shot, Emotional, Expert Mimicry), and confidence scales (1-10, 1-100). We used AUC-ROC, Brier Score, and Expected Calibrati","authors_text":"Ali Soroush, Aref Mahjoub far, Nariman Naderi, Peter R Lewis, Seyed Amir Ahmad Safavi-Naini, Zahra Atf","cross_cats":["cs.AI","cs.CL","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CY","submitted_at":"2025-05-29T17:13:26Z","title":"Evaluating Prompt Engineering Techniques for Accuracy and Confidence Elicitation in Medical LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.00072","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":"sha256:59966167b956ef1f6b073acfdc728c0d6c8f09ec96f6eb9c0575439011077a31","target":"record","created_at":"2026-07-05T11:13:28Z","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":"a91ab13e23b13092085525462d6d046559af57a55f7585d5e8a29dd258de43d7","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CY","submitted_at":"2025-05-29T17:13:26Z","title_canon_sha256":"4859750b1e6dff2bec7874b17228fc3c3270e9d26b2e0ebfe3458d06b8e5de2d"},"schema_version":"1.0","source":{"id":"2506.00072","kind":"arxiv","version":1}},"canonical_sha256":"88c2774dd915b5c37f317e18cead7e04bdf7ee59ad10271c49f99a19d898cf7e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"88c2774dd915b5c37f317e18cead7e04bdf7ee59ad10271c49f99a19d898cf7e","first_computed_at":"2026-07-05T11:13:28.538611Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:13:28.538611Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"UCR/8NvgEWm4m0uxkhamKyzAPPbTcBHtZiCpjrftt436hVFBD0sFcukUb4IX29Dfv2YL1XQEMj+J0egEMHtHCw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:13:28.539187Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.00072","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:59966167b956ef1f6b073acfdc728c0d6c8f09ec96f6eb9c0575439011077a31","sha256:f29b14e65222b2e4f8f5abb9b04f2fd0673a37317e94970177215ab9dfb3bdbd"],"state_sha256":"d7d3424e6ebac26b4cc8fe95669909cd12f2d212b8f36a439f8e2cc5ec5b24f0"}