{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:V43SO6PRDQSMAGI44WDY7X2ERB","short_pith_number":"pith:V43SO6PR","schema_version":"1.0","canonical_sha256":"af372779f11c24c0191ce5878fdf44886864583d8cef81c4936652b966b57bf7","source":{"kind":"arxiv","id":"2402.09199","version":1},"attestation_state":"computed","paper":{"title":"Ten Words Only Still Help: Improving Black-Box AI-Generated Text Detection via Proxy-Guided Efficient Re-Sampling","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Beizhe Hu, Danding Wang, Hao Mi, Juan Cao, Qiang Sheng, Yuhui Shi","submitted_at":"2024-02-14T14:32:16Z","abstract_excerpt":"With the rapidly increasing application of large language models (LLMs), their abuse has caused many undesirable societal problems such as fake news, academic dishonesty, and information pollution. This makes AI-generated text (AIGT) detection of great importance. Among existing methods, white-box methods are generally superior to black-box methods in terms of performance and generalizability, but they require access to LLMs' internal states and are not applicable to black-box settings. In this paper, we propose to estimate word generation probabilities as pseudo white-box features via multipl"},"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":"2402.09199","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-14T14:32:16Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"c1634bcc980e97ef84df393e2715a713a42618c3b1a103f7ed286d24d6bd98dd","abstract_canon_sha256":"e916ecebf9de1768ae25b9786570ef6458b76988fd1af01c674632e113105bfa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:00:01.793781Z","signature_b64":"iVaavhP2I9dXn/4zfFQoUbZJ4+h+gX6JEhTaIUkeVGlhJRE6XtNm3eRPvtnVWMHxxeYwv/3OU0SjKKTn+PtYAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"af372779f11c24c0191ce5878fdf44886864583d8cef81c4936652b966b57bf7","last_reissued_at":"2026-07-05T09:00:01.793288Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:00:01.793288Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Ten Words Only Still Help: Improving Black-Box AI-Generated Text Detection via Proxy-Guided Efficient Re-Sampling","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Beizhe Hu, Danding Wang, Hao Mi, Juan Cao, Qiang Sheng, Yuhui Shi","submitted_at":"2024-02-14T14:32:16Z","abstract_excerpt":"With the rapidly increasing application of large language models (LLMs), their abuse has caused many undesirable societal problems such as fake news, academic dishonesty, and information pollution. This makes AI-generated text (AIGT) detection of great importance. Among existing methods, white-box methods are generally superior to black-box methods in terms of performance and generalizability, but they require access to LLMs' internal states and are not applicable to black-box settings. In this paper, we propose to estimate word generation probabilities as pseudo white-box features via multipl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.09199","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/2402.09199/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":"2402.09199","created_at":"2026-07-05T09:00:01.793355+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.09199v1","created_at":"2026-07-05T09:00:01.793355+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.09199","created_at":"2026-07-05T09:00:01.793355+00:00"},{"alias_kind":"pith_short_12","alias_value":"V43SO6PRDQSM","created_at":"2026-07-05T09:00:01.793355+00:00"},{"alias_kind":"pith_short_16","alias_value":"V43SO6PRDQSMAGI4","created_at":"2026-07-05T09:00:01.793355+00:00"},{"alias_kind":"pith_short_8","alias_value":"V43SO6PR","created_at":"2026-07-05T09:00:01.793355+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2509.22055","citing_title":"RedNote-Vibe: A Dataset for Capturing Temporal Dynamics of AI-Generated Text in Lifestyle Social Media","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/V43SO6PRDQSMAGI44WDY7X2ERB","json":"https://pith.science/pith/V43SO6PRDQSMAGI44WDY7X2ERB.json","graph_json":"https://pith.science/api/pith-number/V43SO6PRDQSMAGI44WDY7X2ERB/graph.json","events_json":"https://pith.science/api/pith-number/V43SO6PRDQSMAGI44WDY7X2ERB/events.json","paper":"https://pith.science/paper/V43SO6PR"},"agent_actions":{"view_html":"https://pith.science/pith/V43SO6PRDQSMAGI44WDY7X2ERB","download_json":"https://pith.science/pith/V43SO6PRDQSMAGI44WDY7X2ERB.json","view_paper":"https://pith.science/paper/V43SO6PR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.09199&json=true","fetch_graph":"https://pith.science/api/pith-number/V43SO6PRDQSMAGI44WDY7X2ERB/graph.json","fetch_events":"https://pith.science/api/pith-number/V43SO6PRDQSMAGI44WDY7X2ERB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V43SO6PRDQSMAGI44WDY7X2ERB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V43SO6PRDQSMAGI44WDY7X2ERB/action/storage_attestation","attest_author":"https://pith.science/pith/V43SO6PRDQSMAGI44WDY7X2ERB/action/author_attestation","sign_citation":"https://pith.science/pith/V43SO6PRDQSMAGI44WDY7X2ERB/action/citation_signature","submit_replication":"https://pith.science/pith/V43SO6PRDQSMAGI44WDY7X2ERB/action/replication_record"}},"created_at":"2026-07-05T09:00:01.793355+00:00","updated_at":"2026-07-05T09:00:01.793355+00:00"}