{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:FTQ6BEYBH4RY7OXDU2KRIB6XZC","short_pith_number":"pith:FTQ6BEYB","canonical_record":{"source":{"id":"2512.14561","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-12-16T16:33:07Z","cross_cats_sorted":[],"title_canon_sha256":"ea32a299f039f107579cacea197bdf684179cd44155ea925726660eff80c492b","abstract_canon_sha256":"2f6ce825818404d6b05f20e15711bf1144e249588f5d4dc672989c2a0925ccca"},"schema_version":"1.0"},"canonical_sha256":"2ce1e093013f238fbae3a6951407d7c8b28b6c777c366ae9b49030eb02c27e6e","source":{"kind":"arxiv","id":"2512.14561","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2512.14561","created_at":"2026-05-27T01:04:53Z"},{"alias_kind":"arxiv_version","alias_value":"2512.14561v2","created_at":"2026-05-27T01:04:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.14561","created_at":"2026-05-27T01:04:53Z"},{"alias_kind":"pith_short_12","alias_value":"FTQ6BEYBH4RY","created_at":"2026-05-27T01:04:53Z"},{"alias_kind":"pith_short_16","alias_value":"FTQ6BEYBH4RY7OXD","created_at":"2026-05-27T01:04:53Z"},{"alias_kind":"pith_short_8","alias_value":"FTQ6BEYB","created_at":"2026-05-27T01:04:53Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:FTQ6BEYBH4RY7OXDU2KRIB6XZC","target":"record","payload":{"canonical_record":{"source":{"id":"2512.14561","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-12-16T16:33:07Z","cross_cats_sorted":[],"title_canon_sha256":"ea32a299f039f107579cacea197bdf684179cd44155ea925726660eff80c492b","abstract_canon_sha256":"2f6ce825818404d6b05f20e15711bf1144e249588f5d4dc672989c2a0925ccca"},"schema_version":"1.0"},"canonical_sha256":"2ce1e093013f238fbae3a6951407d7c8b28b6c777c366ae9b49030eb02c27e6e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T01:04:53.854332Z","signature_b64":"HEkUc2wIsmBv/WXoIRn5nCkv1au3lnkUi8Mk5odw74WxntNIHF+9YE3gNOiPVCfI7t3lzG6Ba3IzV2IUBYffAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2ce1e093013f238fbae3a6951407d7c8b28b6c777c366ae9b49030eb02c27e6e","last_reissued_at":"2026-05-27T01:04:53.853802Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T01:04:53.853802Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2512.14561","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-27T01:04:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1Hs5g1XROP81UoSRa16ZfYjLVIAfWJKizY6wFjc0ZOkZzn5p2C5dMebmoFANfZ0Tzdw6CZjjJoPT7GOAqQYUAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T22:49:35.719384Z"},"content_sha256":"ecf81965d128846e25877770f321b356a2ae09b1dcea9ae695205b4e95e009ca","schema_version":"1.0","event_id":"sha256:ecf81965d128846e25877770f321b356a2ae09b1dcea9ae695205b4e95e009ca"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:FTQ6BEYBH4RY7OXDU2KRIB6XZC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Agreement Between Large Language Models and Human Raters in Essay Scoring: A Research Synthesis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Che Han Chen, Chiho Young-Johnson, Hongli Li, Kevin Fan, Soyoung Lim, Yali Feng","submitted_at":"2025-12-16T16:33:07Z","abstract_excerpt":"Despite the growing promise of large language models (LLMs) in automated essay scoring (AES), empirical findings regarding their reliability compared to human raters remain mixed. Following the PRISMA 2020 guidelines, we synthesized 65 published and unpublished studies from January 2022 to August 2025 that examined agreement between LLM-generated scores and human ratings. Agreement levels varied substantially both across and within studies, with reported values spanning a wide range. Overall, the findings suggest that LLM-human agreement is highly context-dependent. Implications, challenges, a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.14561","kind":"arxiv","version":2},"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/2512.14561/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":"2026-05-27T01:04:53Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Uz5JUB/E1C9fcDW/7OvvQfup2QS8xhMK/Oo/FkxmZoXa4h5tfTc6QEE4WGHLd6jAGhSeBKeiooRdJX9V+oQWAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T22:49:35.719880Z"},"content_sha256":"c2476fccf77bfbec329fa8e66aa1d5b5125476a7fee465b67f0df1a5123e0c50","schema_version":"1.0","event_id":"sha256:c2476fccf77bfbec329fa8e66aa1d5b5125476a7fee465b67f0df1a5123e0c50"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/FTQ6BEYBH4RY7OXDU2KRIB6XZC/bundle.json","state_url":"https://pith.science/pith/FTQ6BEYBH4RY7OXDU2KRIB6XZC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/FTQ6BEYBH4RY7OXDU2KRIB6XZC/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":"2026-08-08T22:49:35Z","links":{"resolver":"https://pith.science/pith/FTQ6BEYBH4RY7OXDU2KRIB6XZC","bundle":"https://pith.science/pith/FTQ6BEYBH4RY7OXDU2KRIB6XZC/bundle.json","state":"https://pith.science/pith/FTQ6BEYBH4RY7OXDU2KRIB6XZC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/FTQ6BEYBH4RY7OXDU2KRIB6XZC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:FTQ6BEYBH4RY7OXDU2KRIB6XZC","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":"2f6ce825818404d6b05f20e15711bf1144e249588f5d4dc672989c2a0925ccca","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-12-16T16:33:07Z","title_canon_sha256":"ea32a299f039f107579cacea197bdf684179cd44155ea925726660eff80c492b"},"schema_version":"1.0","source":{"id":"2512.14561","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2512.14561","created_at":"2026-05-27T01:04:53Z"},{"alias_kind":"arxiv_version","alias_value":"2512.14561v2","created_at":"2026-05-27T01:04:53Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.14561","created_at":"2026-05-27T01:04:53Z"},{"alias_kind":"pith_short_12","alias_value":"FTQ6BEYBH4RY","created_at":"2026-05-27T01:04:53Z"},{"alias_kind":"pith_short_16","alias_value":"FTQ6BEYBH4RY7OXD","created_at":"2026-05-27T01:04:53Z"},{"alias_kind":"pith_short_8","alias_value":"FTQ6BEYB","created_at":"2026-05-27T01:04:53Z"}],"graph_snapshots":[{"event_id":"sha256:c2476fccf77bfbec329fa8e66aa1d5b5125476a7fee465b67f0df1a5123e0c50","target":"graph","created_at":"2026-05-27T01:04:53Z","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/2512.14561/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Despite the growing promise of large language models (LLMs) in automated essay scoring (AES), empirical findings regarding their reliability compared to human raters remain mixed. Following the PRISMA 2020 guidelines, we synthesized 65 published and unpublished studies from January 2022 to August 2025 that examined agreement between LLM-generated scores and human ratings. Agreement levels varied substantially both across and within studies, with reported values spanning a wide range. Overall, the findings suggest that LLM-human agreement is highly context-dependent. Implications, challenges, a","authors_text":"Che Han Chen, Chiho Young-Johnson, Hongli Li, Kevin Fan, Soyoung Lim, Yali Feng","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-12-16T16:33:07Z","title":"Agreement Between Large Language Models and Human Raters in Essay Scoring: A Research Synthesis"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.14561","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:ecf81965d128846e25877770f321b356a2ae09b1dcea9ae695205b4e95e009ca","target":"record","created_at":"2026-05-27T01:04:53Z","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":"2f6ce825818404d6b05f20e15711bf1144e249588f5d4dc672989c2a0925ccca","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-12-16T16:33:07Z","title_canon_sha256":"ea32a299f039f107579cacea197bdf684179cd44155ea925726660eff80c492b"},"schema_version":"1.0","source":{"id":"2512.14561","kind":"arxiv","version":2}},"canonical_sha256":"2ce1e093013f238fbae3a6951407d7c8b28b6c777c366ae9b49030eb02c27e6e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2ce1e093013f238fbae3a6951407d7c8b28b6c777c366ae9b49030eb02c27e6e","first_computed_at":"2026-05-27T01:04:53.853802Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-27T01:04:53.853802Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"HEkUc2wIsmBv/WXoIRn5nCkv1au3lnkUi8Mk5odw74WxntNIHF+9YE3gNOiPVCfI7t3lzG6Ba3IzV2IUBYffAw==","signature_status":"signed_v1","signed_at":"2026-05-27T01:04:53.854332Z","signed_message":"canonical_sha256_bytes"},"source_id":"2512.14561","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ecf81965d128846e25877770f321b356a2ae09b1dcea9ae695205b4e95e009ca","sha256:c2476fccf77bfbec329fa8e66aa1d5b5125476a7fee465b67f0df1a5123e0c50"],"state_sha256":"7c13b11f06954c0e87edca8d730006fc921d0138802f8e87f3ba7d1739134b9d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"h+6LbbcoOKRUSWUU+OH6RyCIZeS6zoJu0NuQMnWJeVsvgnPN/O/SZlgMb2xaWHAS5RNQJmu2BQk0NH54Jk6TBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T22:49:35.723081Z","bundle_sha256":"539b2745092d46988de69dc82e41447e5ab4d9930d570f69319ba69fe943f1ff"}}