{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SX7PTBO5XB6PTT3ZLABYKAAWG2","short_pith_number":"pith:SX7PTBO5","schema_version":"1.0","canonical_sha256":"95fef985ddb87cf9cf79580385001636843dfcab2e746e90e8865cce2ea4fd30","source":{"kind":"arxiv","id":"2507.19219","version":2},"attestation_state":"computed","paper":{"title":"How Much Do Large Language Model Cheat on Evaluation? Benchmarking Overestimation under the One-Time-Pad-Based Framework","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.CL","authors_text":"Haibo Hu, Liantong Yu, Qingqing Ye, Shiyu Zhang, Zi Liang","submitted_at":"2025-07-25T12:39:03Z","abstract_excerpt":"Overestimation in evaluating large language models (LLMs) has become an increasing concern. Due to the contamination of public benchmarks or imbalanced model training, LLMs may achieve unreal evaluation results on public benchmarks, either intentionally or unintentionally, which leads to unfair comparisons among LLMs and undermines their realistic capability assessments. Existing benchmarks attempt to address these issues by keeping test cases permanently secret, mitigating contamination through human evaluation, or repeatedly collecting and constructing new samples. However, these approaches "},"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":"2507.19219","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-07-25T12:39:03Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"25c1af58e34bcf4231b11a4d3b490779913a28a884c4280b6df659625becadde","abstract_canon_sha256":"9ae78d4bf022fd503a679482cbc7b5a316176b41bcdc3211bb8eb76afdb5d54e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:03:13.766279Z","signature_b64":"269JhZN17wc+NpyQrBIn+xAK8g+cEi3J982Evp9vqZMNdKIAB8Sax4f8PHjfVoRwLfRzGt+jjyxkpEffdfFyAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"95fef985ddb87cf9cf79580385001636843dfcab2e746e90e8865cce2ea4fd30","last_reissued_at":"2026-05-26T01:03:13.765708Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:03:13.765708Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How Much Do Large Language Model Cheat on Evaluation? Benchmarking Overestimation under the One-Time-Pad-Based Framework","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.CL","authors_text":"Haibo Hu, Liantong Yu, Qingqing Ye, Shiyu Zhang, Zi Liang","submitted_at":"2025-07-25T12:39:03Z","abstract_excerpt":"Overestimation in evaluating large language models (LLMs) has become an increasing concern. Due to the contamination of public benchmarks or imbalanced model training, LLMs may achieve unreal evaluation results on public benchmarks, either intentionally or unintentionally, which leads to unfair comparisons among LLMs and undermines their realistic capability assessments. Existing benchmarks attempt to address these issues by keeping test cases permanently secret, mitigating contamination through human evaluation, or repeatedly collecting and constructing new samples. However, these approaches "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.19219","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/2507.19219/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":"2507.19219","created_at":"2026-05-26T01:03:13.765780+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.19219v2","created_at":"2026-05-26T01:03:13.765780+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.19219","created_at":"2026-05-26T01:03:13.765780+00:00"},{"alias_kind":"pith_short_12","alias_value":"SX7PTBO5XB6P","created_at":"2026-05-26T01:03:13.765780+00:00"},{"alias_kind":"pith_short_16","alias_value":"SX7PTBO5XB6PTT3Z","created_at":"2026-05-26T01:03:13.765780+00:00"},{"alias_kind":"pith_short_8","alias_value":"SX7PTBO5","created_at":"2026-05-26T01:03:13.765780+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2605.10810","citing_title":"Likelihood scoring for continuations of mathematical text: a self-supervised benchmark with tests for shortcut vulnerabilities","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2605.02442","citing_title":"Measuring AI Reasoning: A Guide for Researchers","ref_index":89,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SX7PTBO5XB6PTT3ZLABYKAAWG2","json":"https://pith.science/pith/SX7PTBO5XB6PTT3ZLABYKAAWG2.json","graph_json":"https://pith.science/api/pith-number/SX7PTBO5XB6PTT3ZLABYKAAWG2/graph.json","events_json":"https://pith.science/api/pith-number/SX7PTBO5XB6PTT3ZLABYKAAWG2/events.json","paper":"https://pith.science/paper/SX7PTBO5"},"agent_actions":{"view_html":"https://pith.science/pith/SX7PTBO5XB6PTT3ZLABYKAAWG2","download_json":"https://pith.science/pith/SX7PTBO5XB6PTT3ZLABYKAAWG2.json","view_paper":"https://pith.science/paper/SX7PTBO5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.19219&json=true","fetch_graph":"https://pith.science/api/pith-number/SX7PTBO5XB6PTT3ZLABYKAAWG2/graph.json","fetch_events":"https://pith.science/api/pith-number/SX7PTBO5XB6PTT3ZLABYKAAWG2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SX7PTBO5XB6PTT3ZLABYKAAWG2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SX7PTBO5XB6PTT3ZLABYKAAWG2/action/storage_attestation","attest_author":"https://pith.science/pith/SX7PTBO5XB6PTT3ZLABYKAAWG2/action/author_attestation","sign_citation":"https://pith.science/pith/SX7PTBO5XB6PTT3ZLABYKAAWG2/action/citation_signature","submit_replication":"https://pith.science/pith/SX7PTBO5XB6PTT3ZLABYKAAWG2/action/replication_record"}},"created_at":"2026-05-26T01:03:13.765780+00:00","updated_at":"2026-05-26T01:03:13.765780+00:00"}