{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:456TNFUFWMF7MFYX37W7FPALST","short_pith_number":"pith:456TNFUF","canonical_record":{"source":{"id":"2406.11657","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-17T15:41:30Z","cross_cats_sorted":["cs.CY"],"title_canon_sha256":"7dd3d8300c31ec75745625e2b31c30248613712e5433aaf1ed2014ec5a879971","abstract_canon_sha256":"4fe89b6d5bb65000ec90d559f4fff60efa6e601787894a96889c7772de64f431"},"schema_version":"1.0"},"canonical_sha256":"e77d369685b30bf61717dfedf2bc0b94d8544179a53f44a17fd80e3b1ea75922","source":{"kind":"arxiv","id":"2406.11657","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.11657","created_at":"2026-07-05T08:32:56Z"},{"alias_kind":"arxiv_version","alias_value":"2406.11657v1","created_at":"2026-07-05T08:32:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.11657","created_at":"2026-07-05T08:32:56Z"},{"alias_kind":"pith_short_12","alias_value":"456TNFUFWMF7","created_at":"2026-07-05T08:32:56Z"},{"alias_kind":"pith_short_16","alias_value":"456TNFUFWMF7MFYX","created_at":"2026-07-05T08:32:56Z"},{"alias_kind":"pith_short_8","alias_value":"456TNFUF","created_at":"2026-07-05T08:32:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:456TNFUFWMF7MFYX37W7FPALST","target":"record","payload":{"canonical_record":{"source":{"id":"2406.11657","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-17T15:41:30Z","cross_cats_sorted":["cs.CY"],"title_canon_sha256":"7dd3d8300c31ec75745625e2b31c30248613712e5433aaf1ed2014ec5a879971","abstract_canon_sha256":"4fe89b6d5bb65000ec90d559f4fff60efa6e601787894a96889c7772de64f431"},"schema_version":"1.0"},"canonical_sha256":"e77d369685b30bf61717dfedf2bc0b94d8544179a53f44a17fd80e3b1ea75922","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:32:56.032436Z","signature_b64":"0uE96A+E2MAo/P1N2wGOzyFv7/82AyH1PYg08xkZgmLlXDKoB6U4FjYGv7TKMIM5fuNf3LvsfDfkg4xWpDLXBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e77d369685b30bf61717dfedf2bc0b94d8544179a53f44a17fd80e3b1ea75922","last_reissued_at":"2026-07-05T08:32:56.031959Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:32:56.031959Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.11657","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":"2026-07-05T08:32:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lPOg8WQ6bWlVB81zm9gZZ6g42G21ePQI2cLM/myAQ7E/XsuAk78QlrxHqtMSjdtpNtFEEB49pdZ96j3cgqklAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T10:21:02.563519Z"},"content_sha256":"8c8158a4b0274c8333e746a9da44c1c941c12af405d00228ba14000041e1833a","schema_version":"1.0","event_id":"sha256:8c8158a4b0274c8333e746a9da44c1c941c12af405d00228ba14000041e1833a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:456TNFUFWMF7MFYX37W7FPALST","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Can LLM be a Personalized Judge?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CY"],"primary_cat":"cs.CL","authors_text":"Nigel Collier, Tiancheng Hu, Yijiang River Dong","submitted_at":"2024-06-17T15:41:30Z","abstract_excerpt":"Ensuring that large language models (LLMs) reflect diverse user values and preferences is crucial as their user bases expand globally. It is therefore encouraging to see the growing interest in LLM personalization within the research community. However, current works often rely on the LLM-as-a-Judge approach for evaluation without thoroughly examining its validity. In this paper, we investigate the reliability of LLM-as-a-Personalized-Judge, asking LLMs to judge user preferences based on personas. Our findings suggest that directly applying LLM-as-a-Personalized-Judge is less reliable than pre"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.11657","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/2406.11657/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":"2026-07-05T08:32:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+Xim8aFUet2VxxT+DimEEivR2hb9TAbojg+Nd52QsxIcsbfr4OYczqc5TksTgFdgWkDN8mugNh1j8Kh+5VQJBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T10:21:02.563894Z"},"content_sha256":"9bb0b82cd0cdceca30befe1ba5708fb47cf4d7d10dad65649eb300d76b8af18d","schema_version":"1.0","event_id":"sha256:9bb0b82cd0cdceca30befe1ba5708fb47cf4d7d10dad65649eb300d76b8af18d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/456TNFUFWMF7MFYX37W7FPALST/bundle.json","state_url":"https://pith.science/pith/456TNFUFWMF7MFYX37W7FPALST/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/456TNFUFWMF7MFYX37W7FPALST/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":"2026-07-07T10:21:02Z","links":{"resolver":"https://pith.science/pith/456TNFUFWMF7MFYX37W7FPALST","bundle":"https://pith.science/pith/456TNFUFWMF7MFYX37W7FPALST/bundle.json","state":"https://pith.science/pith/456TNFUFWMF7MFYX37W7FPALST/state.json","well_known_bundle":"https://pith.science/.well-known/pith/456TNFUFWMF7MFYX37W7FPALST/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:456TNFUFWMF7MFYX37W7FPALST","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":"4fe89b6d5bb65000ec90d559f4fff60efa6e601787894a96889c7772de64f431","cross_cats_sorted":["cs.CY"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-17T15:41:30Z","title_canon_sha256":"7dd3d8300c31ec75745625e2b31c30248613712e5433aaf1ed2014ec5a879971"},"schema_version":"1.0","source":{"id":"2406.11657","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.11657","created_at":"2026-07-05T08:32:56Z"},{"alias_kind":"arxiv_version","alias_value":"2406.11657v1","created_at":"2026-07-05T08:32:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.11657","created_at":"2026-07-05T08:32:56Z"},{"alias_kind":"pith_short_12","alias_value":"456TNFUFWMF7","created_at":"2026-07-05T08:32:56Z"},{"alias_kind":"pith_short_16","alias_value":"456TNFUFWMF7MFYX","created_at":"2026-07-05T08:32:56Z"},{"alias_kind":"pith_short_8","alias_value":"456TNFUF","created_at":"2026-07-05T08:32:56Z"}],"graph_snapshots":[{"event_id":"sha256:9bb0b82cd0cdceca30befe1ba5708fb47cf4d7d10dad65649eb300d76b8af18d","target":"graph","created_at":"2026-07-05T08:32:56Z","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/2406.11657/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Ensuring that large language models (LLMs) reflect diverse user values and preferences is crucial as their user bases expand globally. It is therefore encouraging to see the growing interest in LLM personalization within the research community. However, current works often rely on the LLM-as-a-Judge approach for evaluation without thoroughly examining its validity. In this paper, we investigate the reliability of LLM-as-a-Personalized-Judge, asking LLMs to judge user preferences based on personas. Our findings suggest that directly applying LLM-as-a-Personalized-Judge is less reliable than pre","authors_text":"Nigel Collier, Tiancheng Hu, Yijiang River Dong","cross_cats":["cs.CY"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-17T15:41:30Z","title":"Can LLM be a Personalized Judge?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.11657","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":"sha256:8c8158a4b0274c8333e746a9da44c1c941c12af405d00228ba14000041e1833a","target":"record","created_at":"2026-07-05T08:32:56Z","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":"4fe89b6d5bb65000ec90d559f4fff60efa6e601787894a96889c7772de64f431","cross_cats_sorted":["cs.CY"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-17T15:41:30Z","title_canon_sha256":"7dd3d8300c31ec75745625e2b31c30248613712e5433aaf1ed2014ec5a879971"},"schema_version":"1.0","source":{"id":"2406.11657","kind":"arxiv","version":1}},"canonical_sha256":"e77d369685b30bf61717dfedf2bc0b94d8544179a53f44a17fd80e3b1ea75922","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e77d369685b30bf61717dfedf2bc0b94d8544179a53f44a17fd80e3b1ea75922","first_computed_at":"2026-07-05T08:32:56.031959Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:32:56.031959Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"0uE96A+E2MAo/P1N2wGOzyFv7/82AyH1PYg08xkZgmLlXDKoB6U4FjYGv7TKMIM5fuNf3LvsfDfkg4xWpDLXBg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:32:56.032436Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.11657","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:8c8158a4b0274c8333e746a9da44c1c941c12af405d00228ba14000041e1833a","sha256:9bb0b82cd0cdceca30befe1ba5708fb47cf4d7d10dad65649eb300d76b8af18d"],"state_sha256":"2d7cbca9f66bd2e4d852f88605812a101349a46dc59111600b5a9e5c92591dad"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HScCFS1R1pGkGUE2amaUG9NvqiVh+O2avdUOj4t7DxogZGNu4172Joy1ZqDBgQTQf0y9M4Fvugx9qV47XNHCCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T10:21:02.565920Z","bundle_sha256":"51053f8b5992884ebcb9f78c027a286e2daf2389ef9ad8a1ba24ad2173d475d4"}}