{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4ZM6J2T7ED44VNGIBRDU3MRL7R","short_pith_number":"pith:4ZM6J2T7","schema_version":"1.0","canonical_sha256":"e659e4ea7f20f9cab4c80c474db22bfc621c4c6c2cf927a9ae7e53fe7b29aeab","source":{"kind":"arxiv","id":"2606.16000","version":2},"attestation_state":"computed","paper":{"title":"GRACE-DS: a Guarded Reward-guided Agent Correction Environment in Data Science","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Aleksandr Tsymbalov, Anastasiya Palienko, Artem Epifanov, Danis Zaripov","submitted_at":"2026-06-14T19:58:06Z","abstract_excerpt":"We introduce GRACE-DS, a Guarded Reward-guided Agent Correction Environment in Data Science for pre-deployment evaluation of LLM-powered AutoML agents. GRACE-DS is a set of evaluation metrics in an isolated environment that can be applied to tabular ML tasks specific to a particular organization. It exposes agents to realistic workflow stages, from planning and data inspection through feature engineering, model development, validation, and code repair to final submission, while hidden executable validators measure not only final predictive performance but also leakage avoidance, reproducibilit"},"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":"2606.16000","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-14T19:58:06Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"437aacbd9b2fb9607241b66717fa6a69c64a657a7fddf4cf4e251e1d3939198e","abstract_canon_sha256":"ead18b7c4e88ac5f26ca88a0f30ce43542fa8726f5970d82c8a205f89562028e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:10:56.114128Z","signature_b64":"n2Gf5J0H57xK+GvIJjlmi+UnnADW4XXrXefRC6/epjzmZqFnM3W99A42MNFbvmmmHil7yHDP6YGEbEHhECeXBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e659e4ea7f20f9cab4c80c474db22bfc621c4c6c2cf927a9ae7e53fe7b29aeab","last_reissued_at":"2026-06-19T16:10:56.113658Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:10:56.113658Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"GRACE-DS: a Guarded Reward-guided Agent Correction Environment in Data Science","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Aleksandr Tsymbalov, Anastasiya Palienko, Artem Epifanov, Danis Zaripov","submitted_at":"2026-06-14T19:58:06Z","abstract_excerpt":"We introduce GRACE-DS, a Guarded Reward-guided Agent Correction Environment in Data Science for pre-deployment evaluation of LLM-powered AutoML agents. GRACE-DS is a set of evaluation metrics in an isolated environment that can be applied to tabular ML tasks specific to a particular organization. It exposes agents to realistic workflow stages, from planning and data inspection through feature engineering, model development, validation, and code repair to final submission, while hidden executable validators measure not only final predictive performance but also leakage avoidance, reproducibilit"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.16000","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/2606.16000/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":"2606.16000","created_at":"2026-06-19T16:10:56.113723+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.16000v2","created_at":"2026-06-19T16:10:56.113723+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.16000","created_at":"2026-06-19T16:10:56.113723+00:00"},{"alias_kind":"pith_short_12","alias_value":"4ZM6J2T7ED44","created_at":"2026-06-19T16:10:56.113723+00:00"},{"alias_kind":"pith_short_16","alias_value":"4ZM6J2T7ED44VNGI","created_at":"2026-06-19T16:10:56.113723+00:00"},{"alias_kind":"pith_short_8","alias_value":"4ZM6J2T7","created_at":"2026-06-19T16:10:56.113723+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/4ZM6J2T7ED44VNGIBRDU3MRL7R","json":"https://pith.science/pith/4ZM6J2T7ED44VNGIBRDU3MRL7R.json","graph_json":"https://pith.science/api/pith-number/4ZM6J2T7ED44VNGIBRDU3MRL7R/graph.json","events_json":"https://pith.science/api/pith-number/4ZM6J2T7ED44VNGIBRDU3MRL7R/events.json","paper":"https://pith.science/paper/4ZM6J2T7"},"agent_actions":{"view_html":"https://pith.science/pith/4ZM6J2T7ED44VNGIBRDU3MRL7R","download_json":"https://pith.science/pith/4ZM6J2T7ED44VNGIBRDU3MRL7R.json","view_paper":"https://pith.science/paper/4ZM6J2T7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.16000&json=true","fetch_graph":"https://pith.science/api/pith-number/4ZM6J2T7ED44VNGIBRDU3MRL7R/graph.json","fetch_events":"https://pith.science/api/pith-number/4ZM6J2T7ED44VNGIBRDU3MRL7R/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4ZM6J2T7ED44VNGIBRDU3MRL7R/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4ZM6J2T7ED44VNGIBRDU3MRL7R/action/storage_attestation","attest_author":"https://pith.science/pith/4ZM6J2T7ED44VNGIBRDU3MRL7R/action/author_attestation","sign_citation":"https://pith.science/pith/4ZM6J2T7ED44VNGIBRDU3MRL7R/action/citation_signature","submit_replication":"https://pith.science/pith/4ZM6J2T7ED44VNGIBRDU3MRL7R/action/replication_record"}},"created_at":"2026-06-19T16:10:56.113723+00:00","updated_at":"2026-06-19T16:10:56.113723+00:00"}