{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:DJKUWPK3V6SOBOR6GK2JASRZ4T","short_pith_number":"pith:DJKUWPK3","canonical_record":{"source":{"id":"2606.20227","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-18T13:40:27Z","cross_cats_sorted":["cs.SE"],"title_canon_sha256":"89de1a6d356ee9dc2b71f4b567663214765c9707616ff8967d635ec831996b92","abstract_canon_sha256":"5a27ecbf30e3e49a2a03f06e08ea9551e9a86cc379efa7286a00c24619be3a1d"},"schema_version":"1.0"},"canonical_sha256":"1a554b3d5bafa4e0ba3e32b4904a39e4d9d7d0a044d52c3f7fe7f0f4c8ed73f9","source":{"kind":"arxiv","id":"2606.20227","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.20227","created_at":"2026-06-19T16:13:06Z"},{"alias_kind":"arxiv_version","alias_value":"2606.20227v1","created_at":"2026-06-19T16:13:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.20227","created_at":"2026-06-19T16:13:06Z"},{"alias_kind":"pith_short_12","alias_value":"DJKUWPK3V6SO","created_at":"2026-06-19T16:13:06Z"},{"alias_kind":"pith_short_16","alias_value":"DJKUWPK3V6SOBOR6","created_at":"2026-06-19T16:13:06Z"},{"alias_kind":"pith_short_8","alias_value":"DJKUWPK3","created_at":"2026-06-19T16:13:06Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:DJKUWPK3V6SOBOR6GK2JASRZ4T","target":"record","payload":{"canonical_record":{"source":{"id":"2606.20227","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-18T13:40:27Z","cross_cats_sorted":["cs.SE"],"title_canon_sha256":"89de1a6d356ee9dc2b71f4b567663214765c9707616ff8967d635ec831996b92","abstract_canon_sha256":"5a27ecbf30e3e49a2a03f06e08ea9551e9a86cc379efa7286a00c24619be3a1d"},"schema_version":"1.0"},"canonical_sha256":"1a554b3d5bafa4e0ba3e32b4904a39e4d9d7d0a044d52c3f7fe7f0f4c8ed73f9","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:13:06.383041Z","signature_b64":"eVfxInDhEM+vBj/Yb2gnecfdAJFSvHi/Un5PFfGOQR/g71Ya3KRcykfI5qzt6KATp7us0su84xuEy4LEF+IvAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1a554b3d5bafa4e0ba3e32b4904a39e4d9d7d0a044d52c3f7fe7f0f4c8ed73f9","last_reissued_at":"2026-06-19T16:13:06.382646Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:13:06.382646Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.20227","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-19T16:13:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kroHpyZwRTZI/JSQi0WIJtaT9ehQ/omRxOrKl7XsMa5IHT43zUmHYwfgWq89iWHCvia/nU5T84prM81DEfO3Bg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-28T04:58:30.998178Z"},"content_sha256":"b91bd150923e8003ecbaa63979455afc8ed165e7c3a33fb69effdccc6e6b1bb0","schema_version":"1.0","event_id":"sha256:b91bd150923e8003ecbaa63979455afc8ed165e7c3a33fb69effdccc6e6b1bb0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:DJKUWPK3V6SOBOR6GK2JASRZ4T","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"QMFOL: Benchmarking Large Language Model Reasoning via Quantifiable Monadic First-Order Logic Test Case Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.SE"],"primary_cat":"cs.AI","authors_text":"Kailong Wang, Ling Shi, Lorenz Goette, Tianlong Yu, Xinyi Zheng, Yongxin Zhao","submitted_at":"2026-06-18T13:40:27Z","abstract_excerpt":"Large Language Models (LLMs) have made significant progress in reasoning, particularly in deductive reasoning, which is crucial for high-stakes decision-making. As models improve, evaluation benchmarks should evolve to keep pace. However, existing benchmarks lack fine-grained control over logical complexity and struggle to balance semantic diversity with logical consistency.\n  To address these issues, we propose QMFOL, an automated framework for generating monadic first-order logic reasoning tasks with quantifiable and controllable complexity. It constructs formal logical structures using conj"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.20227","kind":"arxiv","version":1},"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/2606.20227/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":"2026-06-19T16:13:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8rAzWXDkKXNA1GPckCvwzcvZvN35LqV7DbEDhKAEGGZrrwqNw+Yvu2Z1Ndkz5xYcGoczOZQUQctALeJCQBj1Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-28T04:58:30.998560Z"},"content_sha256":"a727d33a52b34198f2f652ead43773cd88ed1adeb12f4d4135e3a40de67e4c0b","schema_version":"1.0","event_id":"sha256:a727d33a52b34198f2f652ead43773cd88ed1adeb12f4d4135e3a40de67e4c0b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DJKUWPK3V6SOBOR6GK2JASRZ4T/bundle.json","state_url":"https://pith.science/pith/DJKUWPK3V6SOBOR6GK2JASRZ4T/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DJKUWPK3V6SOBOR6GK2JASRZ4T/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":"2026-06-28T04:58:30Z","links":{"resolver":"https://pith.science/pith/DJKUWPK3V6SOBOR6GK2JASRZ4T","bundle":"https://pith.science/pith/DJKUWPK3V6SOBOR6GK2JASRZ4T/bundle.json","state":"https://pith.science/pith/DJKUWPK3V6SOBOR6GK2JASRZ4T/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DJKUWPK3V6SOBOR6GK2JASRZ4T/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:DJKUWPK3V6SOBOR6GK2JASRZ4T","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":"5a27ecbf30e3e49a2a03f06e08ea9551e9a86cc379efa7286a00c24619be3a1d","cross_cats_sorted":["cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-18T13:40:27Z","title_canon_sha256":"89de1a6d356ee9dc2b71f4b567663214765c9707616ff8967d635ec831996b92"},"schema_version":"1.0","source":{"id":"2606.20227","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.20227","created_at":"2026-06-19T16:13:06Z"},{"alias_kind":"arxiv_version","alias_value":"2606.20227v1","created_at":"2026-06-19T16:13:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.20227","created_at":"2026-06-19T16:13:06Z"},{"alias_kind":"pith_short_12","alias_value":"DJKUWPK3V6SO","created_at":"2026-06-19T16:13:06Z"},{"alias_kind":"pith_short_16","alias_value":"DJKUWPK3V6SOBOR6","created_at":"2026-06-19T16:13:06Z"},{"alias_kind":"pith_short_8","alias_value":"DJKUWPK3","created_at":"2026-06-19T16:13:06Z"}],"graph_snapshots":[{"event_id":"sha256:a727d33a52b34198f2f652ead43773cd88ed1adeb12f4d4135e3a40de67e4c0b","target":"graph","created_at":"2026-06-19T16:13:06Z","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/2606.20227/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) have made significant progress in reasoning, particularly in deductive reasoning, which is crucial for high-stakes decision-making. As models improve, evaluation benchmarks should evolve to keep pace. However, existing benchmarks lack fine-grained control over logical complexity and struggle to balance semantic diversity with logical consistency.\n  To address these issues, we propose QMFOL, an automated framework for generating monadic first-order logic reasoning tasks with quantifiable and controllable complexity. It constructs formal logical structures using conj","authors_text":"Kailong Wang, Ling Shi, Lorenz Goette, Tianlong Yu, Xinyi Zheng, Yongxin Zhao","cross_cats":["cs.SE"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-18T13:40:27Z","title":"QMFOL: Benchmarking Large Language Model Reasoning via Quantifiable Monadic First-Order Logic Test Case Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.20227","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:b91bd150923e8003ecbaa63979455afc8ed165e7c3a33fb69effdccc6e6b1bb0","target":"record","created_at":"2026-06-19T16:13:06Z","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":"5a27ecbf30e3e49a2a03f06e08ea9551e9a86cc379efa7286a00c24619be3a1d","cross_cats_sorted":["cs.SE"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-18T13:40:27Z","title_canon_sha256":"89de1a6d356ee9dc2b71f4b567663214765c9707616ff8967d635ec831996b92"},"schema_version":"1.0","source":{"id":"2606.20227","kind":"arxiv","version":1}},"canonical_sha256":"1a554b3d5bafa4e0ba3e32b4904a39e4d9d7d0a044d52c3f7fe7f0f4c8ed73f9","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1a554b3d5bafa4e0ba3e32b4904a39e4d9d7d0a044d52c3f7fe7f0f4c8ed73f9","first_computed_at":"2026-06-19T16:13:06.382646Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-19T16:13:06.382646Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"eVfxInDhEM+vBj/Yb2gnecfdAJFSvHi/Un5PFfGOQR/g71Ya3KRcykfI5qzt6KATp7us0su84xuEy4LEF+IvAQ==","signature_status":"signed_v1","signed_at":"2026-06-19T16:13:06.383041Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.20227","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b91bd150923e8003ecbaa63979455afc8ed165e7c3a33fb69effdccc6e6b1bb0","sha256:a727d33a52b34198f2f652ead43773cd88ed1adeb12f4d4135e3a40de67e4c0b"],"state_sha256":"6532705c9775dc73b6cc12640a08172951d3d4e24db192ee59538b2c21ed0d6e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Dx++2IQQcqmDfS5XKBv9m4kBlVZFOejIElgvB/fLgLCHZUBmBFI6+hqb98b0MEavmy8qxRFkbMtHqiGynUKBAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-28T04:58:31.000383Z","bundle_sha256":"082e11873c20d7ab1e718bc54401e7fdeb44e0bf8b63b862319173fac45533f2"}}