{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:FP4HUMRTCMPIXGA277QHKAJE2I","short_pith_number":"pith:FP4HUMRT","canonical_record":{"source":{"id":"2604.01993","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-04-02T12:59:30Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"05a3e1cc8ec5632bc59ec087b24e0058034fa1f14ebb35d698457c12a9c4c456","abstract_canon_sha256":"4e28e56913322b5b6cdf8833390d306df2c82a9c40990a93ea8b69ef94b8ed63"},"schema_version":"1.0"},"canonical_sha256":"2bf87a3233131e8b981affe0750124d23c05b3040d11e65fbac2867229ce6a10","source":{"kind":"arxiv","id":"2604.01993","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.01993","created_at":"2026-06-10T01:11:00Z"},{"alias_kind":"arxiv_version","alias_value":"2604.01993v2","created_at":"2026-06-10T01:11:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.01993","created_at":"2026-06-10T01:11:00Z"},{"alias_kind":"pith_short_12","alias_value":"FP4HUMRTCMPI","created_at":"2026-06-10T01:11:00Z"},{"alias_kind":"pith_short_16","alias_value":"FP4HUMRTCMPIXGA2","created_at":"2026-06-10T01:11:00Z"},{"alias_kind":"pith_short_8","alias_value":"FP4HUMRT","created_at":"2026-06-10T01:11:00Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:FP4HUMRTCMPIXGA277QHKAJE2I","target":"record","payload":{"canonical_record":{"source":{"id":"2604.01993","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-04-02T12:59:30Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"05a3e1cc8ec5632bc59ec087b24e0058034fa1f14ebb35d698457c12a9c4c456","abstract_canon_sha256":"4e28e56913322b5b6cdf8833390d306df2c82a9c40990a93ea8b69ef94b8ed63"},"schema_version":"1.0"},"canonical_sha256":"2bf87a3233131e8b981affe0750124d23c05b3040d11e65fbac2867229ce6a10","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-10T01:11:00.437382Z","signature_b64":"S3w+qIMOJOpgR14KvGr0sETg0k2Wx39QsmCgYPHfbt98tx8SQ3T2oG41KnHTLGwUrrc0vmwzmY0bX52iVKicAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2bf87a3233131e8b981affe0750124d23c05b3040d11e65fbac2867229ce6a10","last_reissued_at":"2026-06-10T01:11:00.436426Z","signature_status":"signed_v1","first_computed_at":"2026-06-10T01:11:00.436426Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2604.01993","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-10T01:11:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uWYFCdTiq+aYl5eAxbOSY4NdG3acOWGWFl94voH/bQPPIl7aBWDz+VtDhXRO0/+k3+/DTpzwbZoXySLenUxlAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-11T20:43:17.902415Z"},"content_sha256":"1a6959776c001458ae066cd1f831a14b61ffdd08a0507894a418b06e382b775a","schema_version":"1.0","event_id":"sha256:1a6959776c001458ae066cd1f831a14b61ffdd08a0507894a418b06e382b775a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:FP4HUMRTCMPIXGA277QHKAJE2I","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"SAFE: An LLM-as-Verifier Framework for Evidence-Grounded Multi-Hop Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Daeyong Kwon, Seung-won Hwang, Soyoung Yoon","submitted_at":"2026-04-02T12:59:30Z","abstract_excerpt":"Multi-hop QA benchmarks often reward Large Language Models (LLMs) for spurious correctness, where models reach correct answers through invalid intermediate reasoning. We propose SAFE, an LLM-as-verifier framework for evidence-grounded multi-hop QA. Rather than judging only the final answer after generation, SAFE verifies reasoning during generation by checking intermediate steps against the provided passages and previous reasoning trajectory. To make this process checkable, SAFE decomposes reasoning into atomic, evidence-grounded units represented with Knowledge Graph (KG) triples. At train-ti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2604.01993","kind":"arxiv","version":2},"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/2604.01993/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-10T01:11:00Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"th14/3RXvh0cetr/s2XTKEN4HgyPqlrq93fjP7lOU+IGVllkonEXfy2VJlBa3DVAWrapsMsft9Y9xoLLz3bEAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-11T20:43:17.903253Z"},"content_sha256":"0a149c7f12206f747b3501cd1c8d0b48e8cf84448b2f8e6452b5b9b2747bebae","schema_version":"1.0","event_id":"sha256:0a149c7f12206f747b3501cd1c8d0b48e8cf84448b2f8e6452b5b9b2747bebae"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FP4HUMRTCMPIXGA277QHKAJE2I/bundle.json","state_url":"https://pith.science/pith/FP4HUMRTCMPIXGA277QHKAJE2I/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FP4HUMRTCMPIXGA277QHKAJE2I/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-11T20:43:17Z","links":{"resolver":"https://pith.science/pith/FP4HUMRTCMPIXGA277QHKAJE2I","bundle":"https://pith.science/pith/FP4HUMRTCMPIXGA277QHKAJE2I/bundle.json","state":"https://pith.science/pith/FP4HUMRTCMPIXGA277QHKAJE2I/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FP4HUMRTCMPIXGA277QHKAJE2I/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:FP4HUMRTCMPIXGA277QHKAJE2I","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":"4e28e56913322b5b6cdf8833390d306df2c82a9c40990a93ea8b69ef94b8ed63","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-04-02T12:59:30Z","title_canon_sha256":"05a3e1cc8ec5632bc59ec087b24e0058034fa1f14ebb35d698457c12a9c4c456"},"schema_version":"1.0","source":{"id":"2604.01993","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.01993","created_at":"2026-06-10T01:11:00Z"},{"alias_kind":"arxiv_version","alias_value":"2604.01993v2","created_at":"2026-06-10T01:11:00Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.01993","created_at":"2026-06-10T01:11:00Z"},{"alias_kind":"pith_short_12","alias_value":"FP4HUMRTCMPI","created_at":"2026-06-10T01:11:00Z"},{"alias_kind":"pith_short_16","alias_value":"FP4HUMRTCMPIXGA2","created_at":"2026-06-10T01:11:00Z"},{"alias_kind":"pith_short_8","alias_value":"FP4HUMRT","created_at":"2026-06-10T01:11:00Z"}],"graph_snapshots":[{"event_id":"sha256:0a149c7f12206f747b3501cd1c8d0b48e8cf84448b2f8e6452b5b9b2747bebae","target":"graph","created_at":"2026-06-10T01:11:00Z","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/2604.01993/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multi-hop QA benchmarks often reward Large Language Models (LLMs) for spurious correctness, where models reach correct answers through invalid intermediate reasoning. We propose SAFE, an LLM-as-verifier framework for evidence-grounded multi-hop QA. Rather than judging only the final answer after generation, SAFE verifies reasoning during generation by checking intermediate steps against the provided passages and previous reasoning trajectory. To make this process checkable, SAFE decomposes reasoning into atomic, evidence-grounded units represented with Knowledge Graph (KG) triples. At train-ti","authors_text":"Daeyong Kwon, Seung-won Hwang, Soyoung Yoon","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-04-02T12:59:30Z","title":"SAFE: An LLM-as-Verifier Framework for Evidence-Grounded Multi-Hop Reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2604.01993","kind":"arxiv","version":2},"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":"sha256:1a6959776c001458ae066cd1f831a14b61ffdd08a0507894a418b06e382b775a","target":"record","created_at":"2026-06-10T01:11:00Z","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":"4e28e56913322b5b6cdf8833390d306df2c82a9c40990a93ea8b69ef94b8ed63","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-04-02T12:59:30Z","title_canon_sha256":"05a3e1cc8ec5632bc59ec087b24e0058034fa1f14ebb35d698457c12a9c4c456"},"schema_version":"1.0","source":{"id":"2604.01993","kind":"arxiv","version":2}},"canonical_sha256":"2bf87a3233131e8b981affe0750124d23c05b3040d11e65fbac2867229ce6a10","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2bf87a3233131e8b981affe0750124d23c05b3040d11e65fbac2867229ce6a10","first_computed_at":"2026-06-10T01:11:00.436426Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-10T01:11:00.436426Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"S3w+qIMOJOpgR14KvGr0sETg0k2Wx39QsmCgYPHfbt98tx8SQ3T2oG41KnHTLGwUrrc0vmwzmY0bX52iVKicAA==","signature_status":"signed_v1","signed_at":"2026-06-10T01:11:00.437382Z","signed_message":"canonical_sha256_bytes"},"source_id":"2604.01993","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1a6959776c001458ae066cd1f831a14b61ffdd08a0507894a418b06e382b775a","sha256:0a149c7f12206f747b3501cd1c8d0b48e8cf84448b2f8e6452b5b9b2747bebae"],"state_sha256":"98e9ef0f8d7c75d75ed13e96d8851f1abe91df682f1e291ca3fc42b3459876b3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Je3gfexbZdJ9NOFmPtWCFlonOP7Lbtie4zt6wH0zHPzpg2By6aSTRRTrKw0b1A2aseYiQoyE66w6pO2if6RQBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-11T20:43:17.907216Z","bundle_sha256":"dbbf1d68f320cd504d94b8f5f3c455fd34293a375ae680a3761329f7aa50379f"}}