{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:CSWOVID2JHYGBXU33HEFFBPR5O","short_pith_number":"pith:CSWOVID2","schema_version":"1.0","canonical_sha256":"14aceaa07a49f060de9bd9c85285f1eb895eeb7ab0a94c14763dccb6d1bb773b","source":{"kind":"arxiv","id":"2410.10876","version":1},"attestation_state":"computed","paper":{"title":"FreqMark: Frequency-Based Watermark for Sentence-Level Detection of LLM-Generated Text","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Kun Zhang, Victor S. Sheng, Zhenyu Xu","submitted_at":"2024-10-09T05:01:48Z","abstract_excerpt":"The increasing use of Large Language Models (LLMs) for generating highly coherent and contextually relevant text introduces new risks, including misuse for unethical purposes such as disinformation or academic dishonesty. To address these challenges, we propose FreqMark, a novel watermarking technique that embeds detectable frequency-based watermarks in LLM-generated text during the token sampling process. The method leverages periodic signals to guide token selection, creating a watermark that can be detected with Short-Time Fourier Transform (STFT) analysis. This approach enables accurate id"},"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":"2410.10876","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-09T05:01:48Z","cross_cats_sorted":["cs.CR","cs.LG"],"title_canon_sha256":"4cb4192396191fc21e3fa052b275febb3f839646566bc240024330c6f38abc33","abstract_canon_sha256":"c2067ee7cc941e50b6b977ab001d320da3440f6c2a50fb8ae8de2e7ba8b4bffd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:20:26.365485Z","signature_b64":"jE30WWzacZp6Nb6VdJEKcIpIUyLi5jl++Eu3vII2oGQhnaKlNJsGK8M8lsE1SF8G1/ct4m9LOqX33f7BzVwVDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"14aceaa07a49f060de9bd9c85285f1eb895eeb7ab0a94c14763dccb6d1bb773b","last_reissued_at":"2026-07-05T09:20:26.365102Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:20:26.365102Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"FreqMark: Frequency-Based Watermark for Sentence-Level Detection of LLM-Generated Text","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Kun Zhang, Victor S. Sheng, Zhenyu Xu","submitted_at":"2024-10-09T05:01:48Z","abstract_excerpt":"The increasing use of Large Language Models (LLMs) for generating highly coherent and contextually relevant text introduces new risks, including misuse for unethical purposes such as disinformation or academic dishonesty. To address these challenges, we propose FreqMark, a novel watermarking technique that embeds detectable frequency-based watermarks in LLM-generated text during the token sampling process. The method leverages periodic signals to guide token selection, creating a watermark that can be detected with Short-Time Fourier Transform (STFT) analysis. This approach enables accurate id"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.10876","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/2410.10876/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":"2410.10876","created_at":"2026-07-05T09:20:26.365159+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.10876v1","created_at":"2026-07-05T09:20:26.365159+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.10876","created_at":"2026-07-05T09:20:26.365159+00:00"},{"alias_kind":"pith_short_12","alias_value":"CSWOVID2JHYG","created_at":"2026-07-05T09:20:26.365159+00:00"},{"alias_kind":"pith_short_16","alias_value":"CSWOVID2JHYGBXU3","created_at":"2026-07-05T09:20:26.365159+00:00"},{"alias_kind":"pith_short_8","alias_value":"CSWOVID2","created_at":"2026-07-05T09:20:26.365159+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.00838","citing_title":"Stylometry recognizes human and LLM-generated texts in short samples","ref_index":83,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CSWOVID2JHYGBXU33HEFFBPR5O","json":"https://pith.science/pith/CSWOVID2JHYGBXU33HEFFBPR5O.json","graph_json":"https://pith.science/api/pith-number/CSWOVID2JHYGBXU33HEFFBPR5O/graph.json","events_json":"https://pith.science/api/pith-number/CSWOVID2JHYGBXU33HEFFBPR5O/events.json","paper":"https://pith.science/paper/CSWOVID2"},"agent_actions":{"view_html":"https://pith.science/pith/CSWOVID2JHYGBXU33HEFFBPR5O","download_json":"https://pith.science/pith/CSWOVID2JHYGBXU33HEFFBPR5O.json","view_paper":"https://pith.science/paper/CSWOVID2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.10876&json=true","fetch_graph":"https://pith.science/api/pith-number/CSWOVID2JHYGBXU33HEFFBPR5O/graph.json","fetch_events":"https://pith.science/api/pith-number/CSWOVID2JHYGBXU33HEFFBPR5O/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CSWOVID2JHYGBXU33HEFFBPR5O/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CSWOVID2JHYGBXU33HEFFBPR5O/action/storage_attestation","attest_author":"https://pith.science/pith/CSWOVID2JHYGBXU33HEFFBPR5O/action/author_attestation","sign_citation":"https://pith.science/pith/CSWOVID2JHYGBXU33HEFFBPR5O/action/citation_signature","submit_replication":"https://pith.science/pith/CSWOVID2JHYGBXU33HEFFBPR5O/action/replication_record"}},"created_at":"2026-07-05T09:20:26.365159+00:00","updated_at":"2026-07-05T09:20:26.365159+00:00"}