{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:FUTCSY4G7NVHHLSX765QHQDBTS","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":"70d12c511d8111326c0f2ef2b0cf57b37a633b18aae213ed2a7a0158c46c1bb6","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-09-05T18:48:34Z","title_canon_sha256":"ee65bc5467095b1744dcb18431de2a15706ac2cb3dddbb5bb9360d89530f5dd3"},"schema_version":"1.0","source":{"id":"2509.05440","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.05440","created_at":"2026-07-05T12:05:47Z"},{"alias_kind":"arxiv_version","alias_value":"2509.05440v1","created_at":"2026-07-05T12:05:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.05440","created_at":"2026-07-05T12:05:47Z"},{"alias_kind":"pith_short_12","alias_value":"FUTCSY4G7NVH","created_at":"2026-07-05T12:05:47Z"},{"alias_kind":"pith_short_16","alias_value":"FUTCSY4G7NVHHLSX","created_at":"2026-07-05T12:05:47Z"},{"alias_kind":"pith_short_8","alias_value":"FUTCSY4G","created_at":"2026-07-05T12:05:47Z"}],"graph_snapshots":[{"event_id":"sha256:5cdd97d7441de34f41ef33deefa2cd85b87dfcef773869e4a35ddd7efb54f398","target":"graph","created_at":"2026-07-05T12:05:47Z","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/2509.05440/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As large-language models have been increasingly used as automatic raters for evaluating free-form content, including document summarization, dialog, and story generation, work has been dedicated to evaluating such models by measuring their correlations with human judgment. For \\textit{sample-level} performance, methods which operate by using pairwise comparisons between machine-generated text perform well but often lack the ability to assign absolute scores to individual summaries, an ability crucial for use cases that require thresholding. In this work, we propose a direct-scoring method whic","authors_text":"Alexander Shelton, Ashton Williamson, Logan Lawrence","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-09-05T18:48:34Z","title":"Direct-Scoring NLG Evaluators Can Use Pairwise Comparisons Too"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.05440","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":"sha256:20950318eb9c061fb792e659c1aba74bed20f987a709bd592118fc307a6bf3a7","target":"record","created_at":"2026-07-05T12:05:47Z","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":"70d12c511d8111326c0f2ef2b0cf57b37a633b18aae213ed2a7a0158c46c1bb6","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-09-05T18:48:34Z","title_canon_sha256":"ee65bc5467095b1744dcb18431de2a15706ac2cb3dddbb5bb9360d89530f5dd3"},"schema_version":"1.0","source":{"id":"2509.05440","kind":"arxiv","version":1}},"canonical_sha256":"2d26296386fb6a73ae57ffbb03c0619c90d585b16fc4bb9d3cc55142e694ccae","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2d26296386fb6a73ae57ffbb03c0619c90d585b16fc4bb9d3cc55142e694ccae","first_computed_at":"2026-07-05T12:05:47.648591Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T12:05:47.648591Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"5wNTuReiDsnpZoBwVPXw3qb/QoH81es1ldLga5l1KwnY+zbVxiFgUQRqjOb478Y24VLdcMrmZz9RHakkbl7aDg==","signature_status":"signed_v1","signed_at":"2026-07-05T12:05:47.649188Z","signed_message":"canonical_sha256_bytes"},"source_id":"2509.05440","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:20950318eb9c061fb792e659c1aba74bed20f987a709bd592118fc307a6bf3a7","sha256:5cdd97d7441de34f41ef33deefa2cd85b87dfcef773869e4a35ddd7efb54f398"],"state_sha256":"1fd8b70a1bfd5a17217b135a95c47efce42831998941c583bd1ffe9c3918d6b8"}