{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:CDYUXVWHIOCRYPTFQPNND27QUP","short_pith_number":"pith:CDYUXVWH","canonical_record":{"source":{"id":"2506.15215","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-18T07:49:13Z","cross_cats_sorted":[],"title_canon_sha256":"b9ef2e00d327aeb96d015f328988caef5eaaaafe2b6ef436aeb7bcfec3c8eff4","abstract_canon_sha256":"46f2bd226526920da7bdb5f7fe9c30756775d1256140ba96bac55501f5485ad6"},"schema_version":"1.0"},"canonical_sha256":"10f14bd6c743851c3e6583dad1ebf0a3f95538a79074cd1c950dc817094bbf43","source":{"kind":"arxiv","id":"2506.15215","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.15215","created_at":"2026-07-05T11:23:39Z"},{"alias_kind":"arxiv_version","alias_value":"2506.15215v1","created_at":"2026-07-05T11:23:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.15215","created_at":"2026-07-05T11:23:39Z"},{"alias_kind":"pith_short_12","alias_value":"CDYUXVWHIOCR","created_at":"2026-07-05T11:23:39Z"},{"alias_kind":"pith_short_16","alias_value":"CDYUXVWHIOCRYPTF","created_at":"2026-07-05T11:23:39Z"},{"alias_kind":"pith_short_8","alias_value":"CDYUXVWH","created_at":"2026-07-05T11:23:39Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:CDYUXVWHIOCRYPTFQPNND27QUP","target":"record","payload":{"canonical_record":{"source":{"id":"2506.15215","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-18T07:49:13Z","cross_cats_sorted":[],"title_canon_sha256":"b9ef2e00d327aeb96d015f328988caef5eaaaafe2b6ef436aeb7bcfec3c8eff4","abstract_canon_sha256":"46f2bd226526920da7bdb5f7fe9c30756775d1256140ba96bac55501f5485ad6"},"schema_version":"1.0"},"canonical_sha256":"10f14bd6c743851c3e6583dad1ebf0a3f95538a79074cd1c950dc817094bbf43","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:23:39.385465Z","signature_b64":"w0hE873KEkYdFUE/sUI6AlWB1FmHoXuNvgORFLTATO0tFzGLH/FAhfSY+nfNmuH+8G3UMroroKAFp66mZJK+Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"10f14bd6c743851c3e6583dad1ebf0a3f95538a79074cd1c950dc817094bbf43","last_reissued_at":"2026-07-05T11:23:39.384945Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:23:39.384945Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2506.15215","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":"2026-07-05T11:23:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"sTxM5AunWRw3zZmx9dpgxxVxt8hjn75yAedkikyKlYRJd3t52nAMyXJhLd7AHA0dcoRf6tP9iPwwO50mTaNEDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T06:31:03.518220Z"},"content_sha256":"88d86ae87307ffc3abcf5a163ae0c4ef921fb7ecef162203fc885e2477c94b02","schema_version":"1.0","event_id":"sha256:88d86ae87307ffc3abcf5a163ae0c4ef921fb7ecef162203fc885e2477c94b02"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:CDYUXVWHIOCRYPTFQPNND27QUP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"MinosEval: Distinguishing Factoid and Non-Factoid for Tailored Open-Ended QA Evaluation with LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guandong Wang, Jie Zhai, Jingping Liu, Qi Ye, Tong Ruan, Yating Wang, Yongqi Fan","submitted_at":"2025-06-18T07:49:13Z","abstract_excerpt":"Open-ended question answering (QA) is a key task for evaluating the capabilities of large language models (LLMs). Compared to closed-ended QA, it demands longer answer statements, more nuanced reasoning processes, and diverse expressions, making refined and interpretable automatic evaluation both crucial and challenging. Traditional metrics like ROUGE and BERTScore struggle to capture semantic similarities due to different patterns between model responses and reference answers. Current LLM-based evaluation approaches, such as pairwise or listwise comparisons of candidate answers, lack intuitiv"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.15215","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/2506.15215/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":"2026-07-05T11:23:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fEFAK2KSMd0PuWmm4gyKarnoJLlpnQD/qiypA8vXMPPFnWL1oVGKHt4zoaFoALs6gWQ91oIqOJnSwAyQhF2VCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T06:31:03.518756Z"},"content_sha256":"cb97d02fc4117f35ac3715d5c7a81e9da8de0c4d3b0a2c2bc9279d9f6872e068","schema_version":"1.0","event_id":"sha256:cb97d02fc4117f35ac3715d5c7a81e9da8de0c4d3b0a2c2bc9279d9f6872e068"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/CDYUXVWHIOCRYPTFQPNND27QUP/bundle.json","state_url":"https://pith.science/pith/CDYUXVWHIOCRYPTFQPNND27QUP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/CDYUXVWHIOCRYPTFQPNND27QUP/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":"2026-08-16T06:31:03Z","links":{"resolver":"https://pith.science/pith/CDYUXVWHIOCRYPTFQPNND27QUP","bundle":"https://pith.science/pith/CDYUXVWHIOCRYPTFQPNND27QUP/bundle.json","state":"https://pith.science/pith/CDYUXVWHIOCRYPTFQPNND27QUP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/CDYUXVWHIOCRYPTFQPNND27QUP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:CDYUXVWHIOCRYPTFQPNND27QUP","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":"46f2bd226526920da7bdb5f7fe9c30756775d1256140ba96bac55501f5485ad6","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-18T07:49:13Z","title_canon_sha256":"b9ef2e00d327aeb96d015f328988caef5eaaaafe2b6ef436aeb7bcfec3c8eff4"},"schema_version":"1.0","source":{"id":"2506.15215","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.15215","created_at":"2026-07-05T11:23:39Z"},{"alias_kind":"arxiv_version","alias_value":"2506.15215v1","created_at":"2026-07-05T11:23:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.15215","created_at":"2026-07-05T11:23:39Z"},{"alias_kind":"pith_short_12","alias_value":"CDYUXVWHIOCR","created_at":"2026-07-05T11:23:39Z"},{"alias_kind":"pith_short_16","alias_value":"CDYUXVWHIOCRYPTF","created_at":"2026-07-05T11:23:39Z"},{"alias_kind":"pith_short_8","alias_value":"CDYUXVWH","created_at":"2026-07-05T11:23:39Z"}],"graph_snapshots":[{"event_id":"sha256:cb97d02fc4117f35ac3715d5c7a81e9da8de0c4d3b0a2c2bc9279d9f6872e068","target":"graph","created_at":"2026-07-05T11:23:39Z","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/2506.15215/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Open-ended question answering (QA) is a key task for evaluating the capabilities of large language models (LLMs). Compared to closed-ended QA, it demands longer answer statements, more nuanced reasoning processes, and diverse expressions, making refined and interpretable automatic evaluation both crucial and challenging. Traditional metrics like ROUGE and BERTScore struggle to capture semantic similarities due to different patterns between model responses and reference answers. Current LLM-based evaluation approaches, such as pairwise or listwise comparisons of candidate answers, lack intuitiv","authors_text":"Guandong Wang, Jie Zhai, Jingping Liu, Qi Ye, Tong Ruan, Yating Wang, Yongqi Fan","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-18T07:49:13Z","title":"MinosEval: Distinguishing Factoid and Non-Factoid for Tailored Open-Ended QA Evaluation with LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.15215","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":"sha256:88d86ae87307ffc3abcf5a163ae0c4ef921fb7ecef162203fc885e2477c94b02","target":"record","created_at":"2026-07-05T11:23:39Z","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":"46f2bd226526920da7bdb5f7fe9c30756775d1256140ba96bac55501f5485ad6","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-18T07:49:13Z","title_canon_sha256":"b9ef2e00d327aeb96d015f328988caef5eaaaafe2b6ef436aeb7bcfec3c8eff4"},"schema_version":"1.0","source":{"id":"2506.15215","kind":"arxiv","version":1}},"canonical_sha256":"10f14bd6c743851c3e6583dad1ebf0a3f95538a79074cd1c950dc817094bbf43","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"10f14bd6c743851c3e6583dad1ebf0a3f95538a79074cd1c950dc817094bbf43","first_computed_at":"2026-07-05T11:23:39.384945Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:23:39.384945Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"w0hE873KEkYdFUE/sUI6AlWB1FmHoXuNvgORFLTATO0tFzGLH/FAhfSY+nfNmuH+8G3UMroroKAFp66mZJK+Bg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:23:39.385465Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.15215","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:88d86ae87307ffc3abcf5a163ae0c4ef921fb7ecef162203fc885e2477c94b02","sha256:cb97d02fc4117f35ac3715d5c7a81e9da8de0c4d3b0a2c2bc9279d9f6872e068"],"state_sha256":"8f34aee5d5b318e4a56ed1657c89512a6e5e20e038f7d73f919364ed7e8714a7"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eKWgMTAMEDTAQGVXBZia/7vvHm8fjee/HuUFYEEtcnpi25HuBd4Zw2K5E5SwvT9VYz1hKlqnASluQAwwVKhxCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-16T06:31:03.523397Z","bundle_sha256":"fa8625040a3f72ab99640ea9b5795d108e831af7bc77024f99fd369e808c281d"}}