{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:QIZ6RJDEDJ6T6AWJRGNUAE26LM","short_pith_number":"pith:QIZ6RJDE","canonical_record":{"source":{"id":"2409.15133","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.IR","submitted_at":"2024-09-23T15:38:12Z","cross_cats_sorted":[],"title_canon_sha256":"3b43c39c28cea89ac44c9dd811906c5f8b7d09d998a8bcd916e9223c4eee9045","abstract_canon_sha256":"42ab38415f8f57ae8e3ff503afaa19331b5be08160d247758d9ccabe45f45419"},"schema_version":"1.0"},"canonical_sha256":"8233e8a4641a7d3f02c9899b40135e5b1dbaba8cc1fc7a2a58215b4a1b31b780","source":{"kind":"arxiv","id":"2409.15133","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.15133","created_at":"2026-07-05T10:39:21Z"},{"alias_kind":"arxiv_version","alias_value":"2409.15133v2","created_at":"2026-07-05T10:39:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.15133","created_at":"2026-07-05T10:39:21Z"},{"alias_kind":"pith_short_12","alias_value":"QIZ6RJDEDJ6T","created_at":"2026-07-05T10:39:21Z"},{"alias_kind":"pith_short_16","alias_value":"QIZ6RJDEDJ6T6AWJ","created_at":"2026-07-05T10:39:21Z"},{"alias_kind":"pith_short_8","alias_value":"QIZ6RJDE","created_at":"2026-07-05T10:39:21Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:QIZ6RJDEDJ6T6AWJRGNUAE26LM","target":"record","payload":{"canonical_record":{"source":{"id":"2409.15133","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.IR","submitted_at":"2024-09-23T15:38:12Z","cross_cats_sorted":[],"title_canon_sha256":"3b43c39c28cea89ac44c9dd811906c5f8b7d09d998a8bcd916e9223c4eee9045","abstract_canon_sha256":"42ab38415f8f57ae8e3ff503afaa19331b5be08160d247758d9ccabe45f45419"},"schema_version":"1.0"},"canonical_sha256":"8233e8a4641a7d3f02c9899b40135e5b1dbaba8cc1fc7a2a58215b4a1b31b780","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:39:21.154713Z","signature_b64":"tFPxtkJXjJlYB78Bfu0r5GWQjf4PHQB5u1sqpLzVQ8F4rvYd18wkglsDJThwS15kOfFuTHzmXjdFm5uIiPZAAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8233e8a4641a7d3f02c9899b40135e5b1dbaba8cc1fc7a2a58215b4a1b31b780","last_reissued_at":"2026-07-05T10:39:21.154223Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:39:21.154223Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2409.15133","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":"2026-07-05T10:39:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"h4qyxgdnoELaku7MkfQeVOYSjksNg5PSu6dinK0/VSK/4DyuPEPHtUprJ0Jqm7KaDVBKv8BAsfqFzEFRZQY8CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T08:46:11.091808Z"},"content_sha256":"75376e9a57243982cf12d3c72dc6d4c5912bf87fbbf60ae5b904821ab931effc","schema_version":"1.0","event_id":"sha256:75376e9a57243982cf12d3c72dc6d4c5912bf87fbbf60ae5b904821ab931effc"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:QIZ6RJDEDJ6T6AWJRGNUAE26LM","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Don't Use LLMs to Make Relevance Judgments","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Ian Soboroff","submitted_at":"2024-09-23T15:38:12Z","abstract_excerpt":"Making the relevance judgments for a TREC-style test collection can be complex and expensive. A typical TREC track usually involves a team of six contractors working for 2-4 weeks. Those contractors need to be trained and monitored. Software has to be written to support recording relevance judgments correctly and efficiently. The recent advent of large language models that produce astoundingly human-like flowing text output in response to a natural language prompt has inspired IR researchers to wonder how those models might be used in the relevance judgment collection process. At the ACM SIGIR"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.15133","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/2409.15133/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":"2026-07-05T10:39:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JsHuB6gr9eNET768eajo8MhJfcuqaYw48/dJf8Kc5b4Btn4B2LJmb8wQWWeGzoAgDFCSbiD9invkKiHeo0dNDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T08:46:11.092167Z"},"content_sha256":"89931e92b743822a9f95aef78a38177940e843cc68b8dbb6589107635206a456","schema_version":"1.0","event_id":"sha256:89931e92b743822a9f95aef78a38177940e843cc68b8dbb6589107635206a456"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/QIZ6RJDEDJ6T6AWJRGNUAE26LM/bundle.json","state_url":"https://pith.science/pith/QIZ6RJDEDJ6T6AWJRGNUAE26LM/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/QIZ6RJDEDJ6T6AWJRGNUAE26LM/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":"2026-07-06T08:46:11Z","links":{"resolver":"https://pith.science/pith/QIZ6RJDEDJ6T6AWJRGNUAE26LM","bundle":"https://pith.science/pith/QIZ6RJDEDJ6T6AWJRGNUAE26LM/bundle.json","state":"https://pith.science/pith/QIZ6RJDEDJ6T6AWJRGNUAE26LM/state.json","well_known_bundle":"https://pith.science/.well-known/pith/QIZ6RJDEDJ6T6AWJRGNUAE26LM/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:QIZ6RJDEDJ6T6AWJRGNUAE26LM","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":"42ab38415f8f57ae8e3ff503afaa19331b5be08160d247758d9ccabe45f45419","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.IR","submitted_at":"2024-09-23T15:38:12Z","title_canon_sha256":"3b43c39c28cea89ac44c9dd811906c5f8b7d09d998a8bcd916e9223c4eee9045"},"schema_version":"1.0","source":{"id":"2409.15133","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.15133","created_at":"2026-07-05T10:39:21Z"},{"alias_kind":"arxiv_version","alias_value":"2409.15133v2","created_at":"2026-07-05T10:39:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.15133","created_at":"2026-07-05T10:39:21Z"},{"alias_kind":"pith_short_12","alias_value":"QIZ6RJDEDJ6T","created_at":"2026-07-05T10:39:21Z"},{"alias_kind":"pith_short_16","alias_value":"QIZ6RJDEDJ6T6AWJ","created_at":"2026-07-05T10:39:21Z"},{"alias_kind":"pith_short_8","alias_value":"QIZ6RJDE","created_at":"2026-07-05T10:39:21Z"}],"graph_snapshots":[{"event_id":"sha256:89931e92b743822a9f95aef78a38177940e843cc68b8dbb6589107635206a456","target":"graph","created_at":"2026-07-05T10:39:21Z","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/2409.15133/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Making the relevance judgments for a TREC-style test collection can be complex and expensive. A typical TREC track usually involves a team of six contractors working for 2-4 weeks. Those contractors need to be trained and monitored. Software has to be written to support recording relevance judgments correctly and efficiently. The recent advent of large language models that produce astoundingly human-like flowing text output in response to a natural language prompt has inspired IR researchers to wonder how those models might be used in the relevance judgment collection process. At the ACM SIGIR","authors_text":"Ian Soboroff","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.IR","submitted_at":"2024-09-23T15:38:12Z","title":"Don't Use LLMs to Make Relevance Judgments"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.15133","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":"sha256:75376e9a57243982cf12d3c72dc6d4c5912bf87fbbf60ae5b904821ab931effc","target":"record","created_at":"2026-07-05T10:39:21Z","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":"42ab38415f8f57ae8e3ff503afaa19331b5be08160d247758d9ccabe45f45419","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.IR","submitted_at":"2024-09-23T15:38:12Z","title_canon_sha256":"3b43c39c28cea89ac44c9dd811906c5f8b7d09d998a8bcd916e9223c4eee9045"},"schema_version":"1.0","source":{"id":"2409.15133","kind":"arxiv","version":2}},"canonical_sha256":"8233e8a4641a7d3f02c9899b40135e5b1dbaba8cc1fc7a2a58215b4a1b31b780","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8233e8a4641a7d3f02c9899b40135e5b1dbaba8cc1fc7a2a58215b4a1b31b780","first_computed_at":"2026-07-05T10:39:21.154223Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:39:21.154223Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"tFPxtkJXjJlYB78Bfu0r5GWQjf4PHQB5u1sqpLzVQ8F4rvYd18wkglsDJThwS15kOfFuTHzmXjdFm5uIiPZAAw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:39:21.154713Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.15133","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:75376e9a57243982cf12d3c72dc6d4c5912bf87fbbf60ae5b904821ab931effc","sha256:89931e92b743822a9f95aef78a38177940e843cc68b8dbb6589107635206a456"],"state_sha256":"4760faea94517712290fad2a550ed98ed9c9fc4edeeb7b5947b7b652469f55aa"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lArIxNqVZ6DZCd5yUXVqnAIZ6kfl2C/fT+bLRSaOh+HgjmUJGwrZL0B9SUM5Vot7/FQaELAEBJ3qqyXgD23UCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T08:46:11.094062Z","bundle_sha256":"78dc2e7ca41f75eb315b4671b727abe269ab79f3971644a2f1809ffa0e5dcfd4"}}