{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4XOVSYZXYIXJEN7VFBVXSXUROO","short_pith_number":"pith:4XOVSYZX","schema_version":"1.0","canonical_sha256":"e5dd596337c22e9237f5286b795e9173b0a1ab02a3019c847409240c9ff0498e","source":{"kind":"arxiv","id":"2606.24839","version":1},"attestation_state":"computed","paper":{"title":"Grading the Grader: Lessons from Evaluating an Agentic Data Analysis System","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.AP"],"primary_cat":"cs.AI","authors_text":"Kai-Tai Hsu, Tian Zheng","submitted_at":"2026-06-23T17:18:28Z","abstract_excerpt":"Agentic data analysis systems produce rich outputs, including code, numerical results, and verbal diagnostics. This makes them more challenging to evaluate than single-turn LLM responses. It is therefore necessary to distinguish genuine disagreement between an agent's output and a ground-truth answer from grading artifacts. We investigate how reliably automated graders assess such a system and what strategies improve grading quality by applying LAMBDA, a multi-agent data-analysis system, on 153 numerical QRData tasks from DSGym. We develop and evaluate a three-layer human-AI grading cascade: s"},"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":"2606.24839","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-23T17:18:28Z","cross_cats_sorted":["stat.AP"],"title_canon_sha256":"094ef9cd0483fbe763bd1cc8b6df72aa03638253441452d1be256da2c09f3a60","abstract_canon_sha256":"6a9aee1a9524ab5d01842f4b067e1a8297eb86bf22bdbf873916f6f1413d4cc5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-24T01:15:43.618788Z","signature_b64":"Ru7whD05zUiL56T3EkdNzjZtKVPLQ0OShpiEAn/Gv3waJBQcuuGQLkmcOsAO76cGhyy2PI06hFv/HlUkA36wDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e5dd596337c22e9237f5286b795e9173b0a1ab02a3019c847409240c9ff0498e","last_reissued_at":"2026-06-24T01:15:43.618421Z","signature_status":"signed_v1","first_computed_at":"2026-06-24T01:15:43.618421Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Grading the Grader: Lessons from Evaluating an Agentic Data Analysis System","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.AP"],"primary_cat":"cs.AI","authors_text":"Kai-Tai Hsu, Tian Zheng","submitted_at":"2026-06-23T17:18:28Z","abstract_excerpt":"Agentic data analysis systems produce rich outputs, including code, numerical results, and verbal diagnostics. This makes them more challenging to evaluate than single-turn LLM responses. It is therefore necessary to distinguish genuine disagreement between an agent's output and a ground-truth answer from grading artifacts. We investigate how reliably automated graders assess such a system and what strategies improve grading quality by applying LAMBDA, a multi-agent data-analysis system, on 153 numerical QRData tasks from DSGym. We develop and evaluate a three-layer human-AI grading cascade: s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.24839","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/2606.24839/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":"2606.24839","created_at":"2026-06-24T01:15:43.618482+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.24839v1","created_at":"2026-06-24T01:15:43.618482+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.24839","created_at":"2026-06-24T01:15:43.618482+00:00"},{"alias_kind":"pith_short_12","alias_value":"4XOVSYZXYIXJ","created_at":"2026-06-24T01:15:43.618482+00:00"},{"alias_kind":"pith_short_16","alias_value":"4XOVSYZXYIXJEN7V","created_at":"2026-06-24T01:15:43.618482+00:00"},{"alias_kind":"pith_short_8","alias_value":"4XOVSYZX","created_at":"2026-06-24T01:15:43.618482+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4XOVSYZXYIXJEN7VFBVXSXUROO","json":"https://pith.science/pith/4XOVSYZXYIXJEN7VFBVXSXUROO.json","graph_json":"https://pith.science/api/pith-number/4XOVSYZXYIXJEN7VFBVXSXUROO/graph.json","events_json":"https://pith.science/api/pith-number/4XOVSYZXYIXJEN7VFBVXSXUROO/events.json","paper":"https://pith.science/paper/4XOVSYZX"},"agent_actions":{"view_html":"https://pith.science/pith/4XOVSYZXYIXJEN7VFBVXSXUROO","download_json":"https://pith.science/pith/4XOVSYZXYIXJEN7VFBVXSXUROO.json","view_paper":"https://pith.science/paper/4XOVSYZX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.24839&json=true","fetch_graph":"https://pith.science/api/pith-number/4XOVSYZXYIXJEN7VFBVXSXUROO/graph.json","fetch_events":"https://pith.science/api/pith-number/4XOVSYZXYIXJEN7VFBVXSXUROO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4XOVSYZXYIXJEN7VFBVXSXUROO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4XOVSYZXYIXJEN7VFBVXSXUROO/action/storage_attestation","attest_author":"https://pith.science/pith/4XOVSYZXYIXJEN7VFBVXSXUROO/action/author_attestation","sign_citation":"https://pith.science/pith/4XOVSYZXYIXJEN7VFBVXSXUROO/action/citation_signature","submit_replication":"https://pith.science/pith/4XOVSYZXYIXJEN7VFBVXSXUROO/action/replication_record"}},"created_at":"2026-06-24T01:15:43.618482+00:00","updated_at":"2026-06-24T01:15:43.618482+00:00"}