{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:5B3FNTLT5TKPCC3X67UNGQL4UJ","short_pith_number":"pith:5B3FNTLT","schema_version":"1.0","canonical_sha256":"e87656cd73ecd4f10b77f7e8d3417ca2679e2e5f5239304fe45b7fd731b26439","source":{"kind":"arxiv","id":"2212.09058","version":1},"attestation_state":"computed","paper":{"title":"BEATs: Audio Pre-Training with Acoustic Tokenizers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Chengyi Wang, Daniel Tompkins, Furu Wei, Sanyuan Chen, Shujie Liu, Yu Wu, Zhuo Chen","submitted_at":"2022-12-18T10:41:55Z","abstract_excerpt":"The massive growth of self-supervised learning (SSL) has been witnessed in language, vision, speech, and audio domains over the past few years. While discrete label prediction is widely adopted for other modalities, the state-of-the-art audio SSL models still employ reconstruction loss for pre-training. Compared with reconstruction loss, semantic-rich discrete label prediction encourages the SSL model to abstract the high-level audio semantics and discard the redundant details as in human perception. However, a semantic-rich acoustic tokenizer for general audio pre-training is usually not stra"},"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":"2212.09058","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2022-12-18T10:41:55Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG","cs.SD"],"title_canon_sha256":"537ab1a08e0422f6cc25f00f3e2254253fefd6373a7eb2b8a42b29b897ba8366","abstract_canon_sha256":"2310c6d11ab7b5b0f71d5e9a6a9e062a3e92e614288683645b99f4a4483b267d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:26:17.544485Z","signature_b64":"iWeWl4zbTJN9/Z41YTO5Es8zaU7l2jUBtg+0XjWnnNNtN2IL4CUynEph+ezp9Tb8gEoFS6N65fP9qyeSstgXBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e87656cd73ecd4f10b77f7e8d3417ca2679e2e5f5239304fe45b7fd731b26439","last_reissued_at":"2026-07-05T05:26:17.543944Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:26:17.543944Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"BEATs: Audio Pre-Training with Acoustic Tokenizers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG","cs.SD"],"primary_cat":"eess.AS","authors_text":"Chengyi Wang, Daniel Tompkins, Furu Wei, Sanyuan Chen, Shujie Liu, Yu Wu, Zhuo Chen","submitted_at":"2022-12-18T10:41:55Z","abstract_excerpt":"The massive growth of self-supervised learning (SSL) has been witnessed in language, vision, speech, and audio domains over the past few years. While discrete label prediction is widely adopted for other modalities, the state-of-the-art audio SSL models still employ reconstruction loss for pre-training. Compared with reconstruction loss, semantic-rich discrete label prediction encourages the SSL model to abstract the high-level audio semantics and discard the redundant details as in human perception. However, a semantic-rich acoustic tokenizer for general audio pre-training is usually not stra"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2212.09058","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/2212.09058/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":"2212.09058","created_at":"2026-07-05T05:26:17.544011+00:00"},{"alias_kind":"arxiv_version","alias_value":"2212.09058v1","created_at":"2026-07-05T05:26:17.544011+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2212.09058","created_at":"2026-07-05T05:26:17.544011+00:00"},{"alias_kind":"pith_short_12","alias_value":"5B3FNTLT5TKP","created_at":"2026-07-05T05:26:17.544011+00:00"},{"alias_kind":"pith_short_16","alias_value":"5B3FNTLT5TKPCC3X","created_at":"2026-07-05T05:26:17.544011+00:00"},{"alias_kind":"pith_short_8","alias_value":"5B3FNTLT","created_at":"2026-07-05T05:26:17.544011+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":15,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.05196","citing_title":"Unified Audio Intelligence Without Regressing on Text Intelligence","ref_index":274,"is_internal_anchor":true},{"citing_arxiv_id":"2607.00387","citing_title":"From Objectives to Applications: Aligning Architectural Biases in Audio Self-Supervised Learning","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03821","citing_title":"Finding Needles in the Haystack: Transductive Active Labeling in Ecology","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03821","citing_title":"Finding Needles in the Haystack: Transductive Active Labeling in Ecology","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26641","citing_title":"OmniRetriever: Any-to-Any Audio-Video-Text Retrieval via Fusion-as-Teacher Distillation","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15929","citing_title":"MUSCAT: MUltilingual, SCientific ConversATion Benchmark","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2507.16632","citing_title":"Step-Audio 2 Technical Report","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13672","citing_title":"SpurAudio: A Benchmark for Studying Shortcut Learning in Few-Shot Audio Classification","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06357","citing_title":"Memory Efficient Full-gradient Attacks (MEFA) Framework for Adversarial Defense Evaluations","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18460","citing_title":"Learning Invariant Modality Representation for Robust Multimodal Learning from a Causal Inference Perspective","ref_index":291,"is_internal_anchor":false},{"citing_arxiv_id":"2501.09747","citing_title":"FAST: Efficient Action Tokenization for Vision-Language-Action Models","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06702","citing_title":"ULTRAS -- Unified Learning of Transformer Representations for Audio and Speech Signals","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.14129","citing_title":"Don't Let the Video Speak: Audio-Contrastive Preference Optimization for Audio-Visual Language Models","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15929","citing_title":"MUSCAT: MUltilingual, SCientific ConversATion Benchmark","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15849","citing_title":"TinyMU: A Compact Audio-Language Model for Music Understanding","ref_index":29,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5B3FNTLT5TKPCC3X67UNGQL4UJ","json":"https://pith.science/pith/5B3FNTLT5TKPCC3X67UNGQL4UJ.json","graph_json":"https://pith.science/api/pith-number/5B3FNTLT5TKPCC3X67UNGQL4UJ/graph.json","events_json":"https://pith.science/api/pith-number/5B3FNTLT5TKPCC3X67UNGQL4UJ/events.json","paper":"https://pith.science/paper/5B3FNTLT"},"agent_actions":{"view_html":"https://pith.science/pith/5B3FNTLT5TKPCC3X67UNGQL4UJ","download_json":"https://pith.science/pith/5B3FNTLT5TKPCC3X67UNGQL4UJ.json","view_paper":"https://pith.science/paper/5B3FNTLT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2212.09058&json=true","fetch_graph":"https://pith.science/api/pith-number/5B3FNTLT5TKPCC3X67UNGQL4UJ/graph.json","fetch_events":"https://pith.science/api/pith-number/5B3FNTLT5TKPCC3X67UNGQL4UJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5B3FNTLT5TKPCC3X67UNGQL4UJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5B3FNTLT5TKPCC3X67UNGQL4UJ/action/storage_attestation","attest_author":"https://pith.science/pith/5B3FNTLT5TKPCC3X67UNGQL4UJ/action/author_attestation","sign_citation":"https://pith.science/pith/5B3FNTLT5TKPCC3X67UNGQL4UJ/action/citation_signature","submit_replication":"https://pith.science/pith/5B3FNTLT5TKPCC3X67UNGQL4UJ/action/replication_record"}},"created_at":"2026-07-05T05:26:17.544011+00:00","updated_at":"2026-07-05T05:26:17.544011+00:00"}