{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:XPUE2RQTCXLDIV2UJDJIIQ2VRD","short_pith_number":"pith:XPUE2RQT","schema_version":"1.0","canonical_sha256":"bbe84d461315d634575448d284435588eb9498826a9abd4f33437bde51d514e8","source":{"kind":"arxiv","id":"2405.13622","version":1},"attestation_state":"computed","paper":{"title":"Automated Evaluation of Retrieval-Augmented Language Models with Task-Specific Exam Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CL","authors_text":"Anoop Deoras, Behrooz Omidvar-Tehrani, Gauthier Guinet, Laurent Callot","submitted_at":"2024-05-22T13:14:11Z","abstract_excerpt":"We propose a new method to measure the task-specific accuracy of Retrieval-Augmented Large Language Models (RAG). Evaluation is performed by scoring the RAG on an automatically-generated synthetic exam composed of multiple choice questions based on the corpus of documents associated with the task. Our method is an automated, cost-efficient, interpretable, and robust strategy to select the optimal components for a RAG system. We leverage Item Response Theory (IRT) to estimate the quality of an exam and its informativeness on task-specific accuracy. IRT also provides a natural way to iteratively"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2405.13622","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-05-22T13:14:11Z","cross_cats_sorted":["cs.IR"],"title_canon_sha256":"d163a1e0ff3ec59b2c70745c05911e1e4aa4550482b38f3af943f103b25348fa","abstract_canon_sha256":"10647c81fd5b19d857a6c067f4d128b14c1e218a9161b478aba2f2ded74b989e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:22:05.176620Z","signature_b64":"Gg8IqVptuS7Ms7YFYTd2sup1WtlrDhmcVedXdBOhhIGqXGWFdxBXyVF+hchww5nsRms/80+mQBhhsMVhNYdzAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bbe84d461315d634575448d284435588eb9498826a9abd4f33437bde51d514e8","last_reissued_at":"2026-07-05T08:22:05.176139Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:22:05.176139Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Automated Evaluation of Retrieval-Augmented Language Models with Task-Specific Exam Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IR"],"primary_cat":"cs.CL","authors_text":"Anoop Deoras, Behrooz Omidvar-Tehrani, Gauthier Guinet, Laurent Callot","submitted_at":"2024-05-22T13:14:11Z","abstract_excerpt":"We propose a new method to measure the task-specific accuracy of Retrieval-Augmented Large Language Models (RAG). Evaluation is performed by scoring the RAG on an automatically-generated synthetic exam composed of multiple choice questions based on the corpus of documents associated with the task. Our method is an automated, cost-efficient, interpretable, and robust strategy to select the optimal components for a RAG system. We leverage Item Response Theory (IRT) to estimate the quality of an exam and its informativeness on task-specific accuracy. IRT also provides a natural way to iteratively"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.13622","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/2405.13622/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2405.13622","created_at":"2026-07-05T08:22:05.176191+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.13622v1","created_at":"2026-07-05T08:22:05.176191+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.13622","created_at":"2026-07-05T08:22:05.176191+00:00"},{"alias_kind":"pith_short_12","alias_value":"XPUE2RQTCXLD","created_at":"2026-07-05T08:22:05.176191+00:00"},{"alias_kind":"pith_short_16","alias_value":"XPUE2RQTCXLDIV2U","created_at":"2026-07-05T08:22:05.176191+00:00"},{"alias_kind":"pith_short_8","alias_value":"XPUE2RQT","created_at":"2026-07-05T08:22:05.176191+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.10291","citing_title":"TimeSeriesExamAgent: Creating Time Series Reasoning Benchmarks at Scale","ref_index":15,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XPUE2RQTCXLDIV2UJDJIIQ2VRD","json":"https://pith.science/pith/XPUE2RQTCXLDIV2UJDJIIQ2VRD.json","graph_json":"https://pith.science/api/pith-number/XPUE2RQTCXLDIV2UJDJIIQ2VRD/graph.json","events_json":"https://pith.science/api/pith-number/XPUE2RQTCXLDIV2UJDJIIQ2VRD/events.json","paper":"https://pith.science/paper/XPUE2RQT"},"agent_actions":{"view_html":"https://pith.science/pith/XPUE2RQTCXLDIV2UJDJIIQ2VRD","download_json":"https://pith.science/pith/XPUE2RQTCXLDIV2UJDJIIQ2VRD.json","view_paper":"https://pith.science/paper/XPUE2RQT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.13622&json=true","fetch_graph":"https://pith.science/api/pith-number/XPUE2RQTCXLDIV2UJDJIIQ2VRD/graph.json","fetch_events":"https://pith.science/api/pith-number/XPUE2RQTCXLDIV2UJDJIIQ2VRD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XPUE2RQTCXLDIV2UJDJIIQ2VRD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XPUE2RQTCXLDIV2UJDJIIQ2VRD/action/storage_attestation","attest_author":"https://pith.science/pith/XPUE2RQTCXLDIV2UJDJIIQ2VRD/action/author_attestation","sign_citation":"https://pith.science/pith/XPUE2RQTCXLDIV2UJDJIIQ2VRD/action/citation_signature","submit_replication":"https://pith.science/pith/XPUE2RQTCXLDIV2UJDJIIQ2VRD/action/replication_record"}},"created_at":"2026-07-05T08:22:05.176191+00:00","updated_at":"2026-07-05T08:22:05.176191+00:00"}