{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SN7XSC6RTJK4XLAIM5WJH4M2CV","short_pith_number":"pith:SN7XSC6R","schema_version":"1.0","canonical_sha256":"937f790bd19a55cbac08676c93f19a1575622f151d88ea4941a47c35de86c07a","source":{"kind":"arxiv","id":"2507.17747","version":2},"attestation_state":"computed","paper":{"title":"Pretraining on the Test Set Is No Longer All You Need: A Debate-Driven Approach to QA Benchmarks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jinman Zhao, Linbo Cao","submitted_at":"2025-07-23T17:58:14Z","abstract_excerpt":"As frontier language models increasingly saturate standard QA benchmarks, concerns about data contamination, memorization, and escalating dataset creation costs persist. We propose a debate-driven evaluation paradigm that transforms any existing QA dataset into structured adversarial debates--where one model is given the official answer to defend, and another constructs and defends an alternative answer--adjudicated by a judge model blind to the correct solution. By forcing multi-round argumentation, this approach substantially increases difficulty while penalizing shallow memorization, yet re"},"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":"2507.17747","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-07-23T17:58:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"19939f381c16638e2e3827aa59e2a520f38f5a70ac2065ea55a1fd33a91a37aa","abstract_canon_sha256":"cc637abaa082bcf28faf63926252d799a3185cd54eb525f010112a3f3c8ef532"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:50:38.801665Z","signature_b64":"SwRb4a60zzO2PhzZs7WlCOk7JHZQK3G49kX4GTHyU7jTeX8M8OkypIQi4B6pF4B8hj3s3KXcXvQ3Lc68EYTAAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"937f790bd19a55cbac08676c93f19a1575622f151d88ea4941a47c35de86c07a","last_reissued_at":"2026-07-05T11:50:38.801164Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:50:38.801164Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Pretraining on the Test Set Is No Longer All You Need: A Debate-Driven Approach to QA Benchmarks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jinman Zhao, Linbo Cao","submitted_at":"2025-07-23T17:58:14Z","abstract_excerpt":"As frontier language models increasingly saturate standard QA benchmarks, concerns about data contamination, memorization, and escalating dataset creation costs persist. We propose a debate-driven evaluation paradigm that transforms any existing QA dataset into structured adversarial debates--where one model is given the official answer to defend, and another constructs and defends an alternative answer--adjudicated by a judge model blind to the correct solution. By forcing multi-round argumentation, this approach substantially increases difficulty while penalizing shallow memorization, yet re"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.17747","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/2507.17747/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":"2507.17747","created_at":"2026-07-05T11:50:38.801238+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.17747v2","created_at":"2026-07-05T11:50:38.801238+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.17747","created_at":"2026-07-05T11:50:38.801238+00:00"},{"alias_kind":"pith_short_12","alias_value":"SN7XSC6RTJK4","created_at":"2026-07-05T11:50:38.801238+00:00"},{"alias_kind":"pith_short_16","alias_value":"SN7XSC6RTJK4XLAI","created_at":"2026-07-05T11:50:38.801238+00:00"},{"alias_kind":"pith_short_8","alias_value":"SN7XSC6R","created_at":"2026-07-05T11:50:38.801238+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/SN7XSC6RTJK4XLAIM5WJH4M2CV","json":"https://pith.science/pith/SN7XSC6RTJK4XLAIM5WJH4M2CV.json","graph_json":"https://pith.science/api/pith-number/SN7XSC6RTJK4XLAIM5WJH4M2CV/graph.json","events_json":"https://pith.science/api/pith-number/SN7XSC6RTJK4XLAIM5WJH4M2CV/events.json","paper":"https://pith.science/paper/SN7XSC6R"},"agent_actions":{"view_html":"https://pith.science/pith/SN7XSC6RTJK4XLAIM5WJH4M2CV","download_json":"https://pith.science/pith/SN7XSC6RTJK4XLAIM5WJH4M2CV.json","view_paper":"https://pith.science/paper/SN7XSC6R","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.17747&json=true","fetch_graph":"https://pith.science/api/pith-number/SN7XSC6RTJK4XLAIM5WJH4M2CV/graph.json","fetch_events":"https://pith.science/api/pith-number/SN7XSC6RTJK4XLAIM5WJH4M2CV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SN7XSC6RTJK4XLAIM5WJH4M2CV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SN7XSC6RTJK4XLAIM5WJH4M2CV/action/storage_attestation","attest_author":"https://pith.science/pith/SN7XSC6RTJK4XLAIM5WJH4M2CV/action/author_attestation","sign_citation":"https://pith.science/pith/SN7XSC6RTJK4XLAIM5WJH4M2CV/action/citation_signature","submit_replication":"https://pith.science/pith/SN7XSC6RTJK4XLAIM5WJH4M2CV/action/replication_record"}},"created_at":"2026-07-05T11:50:38.801238+00:00","updated_at":"2026-07-05T11:50:38.801238+00:00"}