{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:673VRSFZ2EUW2ZLNTZI3NMBN6G","short_pith_number":"pith:673VRSFZ","canonical_record":{"source":{"id":"2310.14542","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-23T03:48:24Z","cross_cats_sorted":[],"title_canon_sha256":"54d3b93255af95b83b6ab05d811f888ff9973d1ad2fb6236e5747c8367e7b6f6","abstract_canon_sha256":"0188d9d5373dd5b3379d73e6a52665f600c3a2053baa9c4227813b41e075ead5"},"schema_version":"1.0"},"canonical_sha256":"f7f758c8b9d1296d656d9e51b6b02df19c6c60f9fdda8f5149b7846d5ef0f594","source":{"kind":"arxiv","id":"2310.14542","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.14542","created_at":"2026-07-05T07:03:46Z"},{"alias_kind":"arxiv_version","alias_value":"2310.14542v1","created_at":"2026-07-05T07:03:46Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.14542","created_at":"2026-07-05T07:03:46Z"},{"alias_kind":"pith_short_12","alias_value":"673VRSFZ2EUW","created_at":"2026-07-05T07:03:46Z"},{"alias_kind":"pith_short_16","alias_value":"673VRSFZ2EUW2ZLN","created_at":"2026-07-05T07:03:46Z"},{"alias_kind":"pith_short_8","alias_value":"673VRSFZ","created_at":"2026-07-05T07:03:46Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:673VRSFZ2EUW2ZLNTZI3NMBN6G","target":"record","payload":{"canonical_record":{"source":{"id":"2310.14542","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-23T03:48:24Z","cross_cats_sorted":[],"title_canon_sha256":"54d3b93255af95b83b6ab05d811f888ff9973d1ad2fb6236e5747c8367e7b6f6","abstract_canon_sha256":"0188d9d5373dd5b3379d73e6a52665f600c3a2053baa9c4227813b41e075ead5"},"schema_version":"1.0"},"canonical_sha256":"f7f758c8b9d1296d656d9e51b6b02df19c6c60f9fdda8f5149b7846d5ef0f594","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:03:46.158855Z","signature_b64":"tsVsNagy4Yj+YI/NKzh713QijwHmfEsAzxQz5MzlCPQBzhvXwW5DixQxoYuS4ltK+fQTTIVkUFca/g+rxUFMBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f7f758c8b9d1296d656d9e51b6b02df19c6c60f9fdda8f5149b7846d5ef0f594","last_reissued_at":"2026-07-05T07:03:46.158356Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:03:46.158356Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2310.14542","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":"2026-07-05T07:03:46Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iiSwdxd4PbuI7OA8ddw3zTP4bYK7Ysc0EwWatPyre/LQz8dXdffhnVVcFeiYTI6sMRdcirLiYavCF+tfrngABw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-18T13:15:37.057544Z"},"content_sha256":"5f100a23bd2f2126f66713d9aa2592b24bed43fe7a740829372bad42a62a597e","schema_version":"1.0","event_id":"sha256:5f100a23bd2f2126f66713d9aa2592b24bed43fe7a740829372bad42a62a597e"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:673VRSFZ2EUW2ZLNTZI3NMBN6G","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Evaluating Large Language Models on Controlled Generation Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jiao Sun, John Frederick Wieting, Nan Xu, Nanyun Peng, Qian Hu, Rahul Gupta, Wangchunshu Zhou, Xuezhe Ma, Yufei Tian","submitted_at":"2023-10-23T03:48:24Z","abstract_excerpt":"While recent studies have looked into the abilities of large language models in various benchmark tasks, including question generation, reading comprehension, multilingual and etc, there have been few studies looking into the controllability of large language models on generation tasks. We present an extensive analysis of various benchmarks including a sentence planning benchmark with different granularities. After comparing large language models against state-of-the-start finetuned smaller models, we present a spectrum showing large language models falling behind, are comparable, or exceed th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.14542","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/2310.14542/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":"2026-07-05T07:03:46Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PU2jcvJSYqpBVELtO8D412aBz1WmPD4LK9B0AUhm/JVKDI0FLYnsfqcN8+TDmvp2Ko/TfJZmh684oVL7VTJlDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-18T13:15:37.057916Z"},"content_sha256":"0c89767653133b6106ae5cbe5a2afe0a50110e8e1c95546777ea1ccf9aa73617","schema_version":"1.0","event_id":"sha256:0c89767653133b6106ae5cbe5a2afe0a50110e8e1c95546777ea1ccf9aa73617"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/673VRSFZ2EUW2ZLNTZI3NMBN6G/bundle.json","state_url":"https://pith.science/pith/673VRSFZ2EUW2ZLNTZI3NMBN6G/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/673VRSFZ2EUW2ZLNTZI3NMBN6G/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":"2026-07-18T13:15:37Z","links":{"resolver":"https://pith.science/pith/673VRSFZ2EUW2ZLNTZI3NMBN6G","bundle":"https://pith.science/pith/673VRSFZ2EUW2ZLNTZI3NMBN6G/bundle.json","state":"https://pith.science/pith/673VRSFZ2EUW2ZLNTZI3NMBN6G/state.json","well_known_bundle":"https://pith.science/.well-known/pith/673VRSFZ2EUW2ZLNTZI3NMBN6G/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:673VRSFZ2EUW2ZLNTZI3NMBN6G","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":"0188d9d5373dd5b3379d73e6a52665f600c3a2053baa9c4227813b41e075ead5","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-23T03:48:24Z","title_canon_sha256":"54d3b93255af95b83b6ab05d811f888ff9973d1ad2fb6236e5747c8367e7b6f6"},"schema_version":"1.0","source":{"id":"2310.14542","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.14542","created_at":"2026-07-05T07:03:46Z"},{"alias_kind":"arxiv_version","alias_value":"2310.14542v1","created_at":"2026-07-05T07:03:46Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.14542","created_at":"2026-07-05T07:03:46Z"},{"alias_kind":"pith_short_12","alias_value":"673VRSFZ2EUW","created_at":"2026-07-05T07:03:46Z"},{"alias_kind":"pith_short_16","alias_value":"673VRSFZ2EUW2ZLN","created_at":"2026-07-05T07:03:46Z"},{"alias_kind":"pith_short_8","alias_value":"673VRSFZ","created_at":"2026-07-05T07:03:46Z"}],"graph_snapshots":[{"event_id":"sha256:0c89767653133b6106ae5cbe5a2afe0a50110e8e1c95546777ea1ccf9aa73617","target":"graph","created_at":"2026-07-05T07:03:46Z","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/2310.14542/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"While recent studies have looked into the abilities of large language models in various benchmark tasks, including question generation, reading comprehension, multilingual and etc, there have been few studies looking into the controllability of large language models on generation tasks. We present an extensive analysis of various benchmarks including a sentence planning benchmark with different granularities. After comparing large language models against state-of-the-start finetuned smaller models, we present a spectrum showing large language models falling behind, are comparable, or exceed th","authors_text":"Jiao Sun, John Frederick Wieting, Nan Xu, Nanyun Peng, Qian Hu, Rahul Gupta, Wangchunshu Zhou, Xuezhe Ma, Yufei Tian","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-23T03:48:24Z","title":"Evaluating Large Language Models on Controlled Generation Tasks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.14542","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":"sha256:5f100a23bd2f2126f66713d9aa2592b24bed43fe7a740829372bad42a62a597e","target":"record","created_at":"2026-07-05T07:03:46Z","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":"0188d9d5373dd5b3379d73e6a52665f600c3a2053baa9c4227813b41e075ead5","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-23T03:48:24Z","title_canon_sha256":"54d3b93255af95b83b6ab05d811f888ff9973d1ad2fb6236e5747c8367e7b6f6"},"schema_version":"1.0","source":{"id":"2310.14542","kind":"arxiv","version":1}},"canonical_sha256":"f7f758c8b9d1296d656d9e51b6b02df19c6c60f9fdda8f5149b7846d5ef0f594","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f7f758c8b9d1296d656d9e51b6b02df19c6c60f9fdda8f5149b7846d5ef0f594","first_computed_at":"2026-07-05T07:03:46.158356Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:03:46.158356Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"tsVsNagy4Yj+YI/NKzh713QijwHmfEsAzxQz5MzlCPQBzhvXwW5DixQxoYuS4ltK+fQTTIVkUFca/g+rxUFMBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T07:03:46.158855Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.14542","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5f100a23bd2f2126f66713d9aa2592b24bed43fe7a740829372bad42a62a597e","sha256:0c89767653133b6106ae5cbe5a2afe0a50110e8e1c95546777ea1ccf9aa73617"],"state_sha256":"69c008f0a5b463bde5b80783e8f8e649f6bad63af90467ff181a3381891a2f19"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aVoZjxbb1YGY96IoJL5eyAFt4YIXA1oD+ua44AVvF4yOtCAxpiRLTKpqUE8EP0HFYitAR2ydNh4qF41GSTR5Cg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-18T13:15:37.060013Z","bundle_sha256":"1b8c4fa8a21c40eb5b91b5afe60496a950b3af4fbc4a0c96d7248280675db66f"}}