{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:V5A6TIP3SEF5TNYEE32TIATEGD","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":"9c9cd1d10148ed1947c94edb5b0cbdab1f5cf9de9af0c074be8aa07b76561e3f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-18T03:40:11Z","title_canon_sha256":"69b1b019a262f2af2f389761fb57d980ac0c2b9c4e23773dc6c51519c408dabd"},"schema_version":"1.0","source":{"id":"2504.13439","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.13439","created_at":"2026-07-05T11:20:52Z"},{"alias_kind":"arxiv_version","alias_value":"2504.13439v2","created_at":"2026-07-05T11:20:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.13439","created_at":"2026-07-05T11:20:52Z"},{"alias_kind":"pith_short_12","alias_value":"V5A6TIP3SEF5","created_at":"2026-07-05T11:20:52Z"},{"alias_kind":"pith_short_16","alias_value":"V5A6TIP3SEF5TNYE","created_at":"2026-07-05T11:20:52Z"},{"alias_kind":"pith_short_8","alias_value":"V5A6TIP3","created_at":"2026-07-05T11:20:52Z"}],"graph_snapshots":[{"event_id":"sha256:e36704c1b4b685b53014d965c99cd9baab7dca7d77e6127bda99e746611691e1","target":"graph","created_at":"2026-07-05T11:20:52Z","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/2504.13439/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Evaluating generative models with open-ended generation is challenging due to inconsistencies in response formats. Multiple-choice (MC) evaluation mitigates this issue, but generating high-quality distractors is time-consuming and labor-intensive. We introduce D-GEN, the first open-source distractor generator model that transforms open-ended data into an MC format. To evaluate distractor quality, we propose two novel methods: (1) ranking alignment, ensuring generated distractors retain the discriminatory power of ground-truth distractors, and (2) entropy analysis, comparing model confidence di","authors_text":"Grace Byun, Jinho D. Choi","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-18T03:40:11Z","title":"D-GEN: Automatic Distractor Generation and Evaluation for Reliable Assessment of Generative Model"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.13439","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":"sha256:662ec7a7c03d8ac6aece56e795aab5c53ec39e7f9d9443e75c1f2528e2fc475d","target":"record","created_at":"2026-07-05T11:20:52Z","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":"9c9cd1d10148ed1947c94edb5b0cbdab1f5cf9de9af0c074be8aa07b76561e3f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-04-18T03:40:11Z","title_canon_sha256":"69b1b019a262f2af2f389761fb57d980ac0c2b9c4e23773dc6c51519c408dabd"},"schema_version":"1.0","source":{"id":"2504.13439","kind":"arxiv","version":2}},"canonical_sha256":"af41e9a1fb910bd9b70426f534026430c2f46ea6a8d08342399aa0ab8a50c251","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"af41e9a1fb910bd9b70426f534026430c2f46ea6a8d08342399aa0ab8a50c251","first_computed_at":"2026-07-05T11:20:52.843337Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:20:52.843337Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZsocnvfgFatBL/YmuvoiDXzv/WgXaBQVmc0ZpM8p6HLoj/1rcUFcyEZ24wjRAW9FjGiZcBSERAYIEsTjvOOUAA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:20:52.843828Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.13439","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:662ec7a7c03d8ac6aece56e795aab5c53ec39e7f9d9443e75c1f2528e2fc475d","sha256:e36704c1b4b685b53014d965c99cd9baab7dca7d77e6127bda99e746611691e1"],"state_sha256":"ad8716126d723d9180fe2aa4ab0f97c94f5787583eec7fe8885dfff3f4f8d7a7"}