{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NY2AY2XCSU2JC6IDDCQW5DALM5","short_pith_number":"pith:NY2AY2XC","schema_version":"1.0","canonical_sha256":"6e340c6ae2953491790318a16e8c0b6777ad6b67d2cfbf0f726572e8bf928da5","source":{"kind":"arxiv","id":"2502.15637","version":2},"attestation_state":"computed","paper":{"title":"Mantis: Lightweight Foundation Model for Time Series Classification","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Hongbo Guo, Ievgen Redko, Jianfeng Zhang, Lujia Pan, Malik Tiomoko, Marius Alonso, Quentin Bouniot, Romain Ilbert, Shifeng Xie, Simon Roschmann, Songkang Wen, Vasilii Feofanov, Zeynep Akata","submitted_at":"2025-02-21T18:06:09Z","abstract_excerpt":"While foundation models have revolutionized various domains, their application to time series classification remains rather under-explored, with existing literature predominantly focused on forecasting. To bridge this gap, we introduce \\textbf{Mantis}, a transformer-based foundation model pre-trained exclusively on synthetic data via self-supervised contrastive learning. We demonstrate that effective tokenization is critical to unlocking the full potential of transformers, proposing a novel token generator unit. Furthermore, we introduce an enhanced test-time methodology that bridges the perfo"},"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":"2502.15637","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-21T18:06:09Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"0084ac50abc8aca404179dca59d8c40487b5a7e35439ee75f7010e634fb4dbcc","abstract_canon_sha256":"f85723efffde6a53dff5dac86af508e06768e774c6f026319af161253b508c3c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-01T00:17:07.882298Z","signature_b64":"JZF9Kv+M5bgiVOTy0e4ATdyya03MjcF7vMZBTL0PtZY+25jvsDusiWR2I6d6Hn7ZpeU2qlsxngjZWgXszO1IAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6e340c6ae2953491790318a16e8c0b6777ad6b67d2cfbf0f726572e8bf928da5","last_reissued_at":"2026-07-01T00:17:07.881808Z","signature_status":"signed_v1","first_computed_at":"2026-07-01T00:17:07.881808Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mantis: Lightweight Foundation Model for Time Series Classification","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Hongbo Guo, Ievgen Redko, Jianfeng Zhang, Lujia Pan, Malik Tiomoko, Marius Alonso, Quentin Bouniot, Romain Ilbert, Shifeng Xie, Simon Roschmann, Songkang Wen, Vasilii Feofanov, Zeynep Akata","submitted_at":"2025-02-21T18:06:09Z","abstract_excerpt":"While foundation models have revolutionized various domains, their application to time series classification remains rather under-explored, with existing literature predominantly focused on forecasting. To bridge this gap, we introduce \\textbf{Mantis}, a transformer-based foundation model pre-trained exclusively on synthetic data via self-supervised contrastive learning. We demonstrate that effective tokenization is critical to unlocking the full potential of transformers, proposing a novel token generator unit. Furthermore, we introduce an enhanced test-time methodology that bridges the perfo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.15637","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/2502.15637/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":"2502.15637","created_at":"2026-07-01T00:17:07.881868+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.15637v2","created_at":"2026-07-01T00:17:07.881868+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.15637","created_at":"2026-07-01T00:17:07.881868+00:00"},{"alias_kind":"pith_short_12","alias_value":"NY2AY2XCSU2J","created_at":"2026-07-01T00:17:07.881868+00:00"},{"alias_kind":"pith_short_16","alias_value":"NY2AY2XCSU2JC6ID","created_at":"2026-07-01T00:17:07.881868+00:00"},{"alias_kind":"pith_short_8","alias_value":"NY2AY2XC","created_at":"2026-07-01T00:17:07.881868+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":4,"sample":[{"citing_arxiv_id":"2606.30410","citing_title":"Beyond IID: How General Are Tabular Foundation Models, Really?","ref_index":157,"is_internal_anchor":true},{"citing_arxiv_id":"2503.07259","citing_title":"COMODO: Cross-Modal Video-to-IMU Distillation for Efficient Egocentric Human Activity Recognition","ref_index":11,"is_internal_anchor":true},{"citing_arxiv_id":"2510.06063","citing_title":"TelecomTS: A Multi-Modal Observability Dataset for Time Series and Language Analysis","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08117","citing_title":"Modular Retrieval-Augmented Generalization for Human Action Recognition","ref_index":24,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NY2AY2XCSU2JC6IDDCQW5DALM5","json":"https://pith.science/pith/NY2AY2XCSU2JC6IDDCQW5DALM5.json","graph_json":"https://pith.science/api/pith-number/NY2AY2XCSU2JC6IDDCQW5DALM5/graph.json","events_json":"https://pith.science/api/pith-number/NY2AY2XCSU2JC6IDDCQW5DALM5/events.json","paper":"https://pith.science/paper/NY2AY2XC"},"agent_actions":{"view_html":"https://pith.science/pith/NY2AY2XCSU2JC6IDDCQW5DALM5","download_json":"https://pith.science/pith/NY2AY2XCSU2JC6IDDCQW5DALM5.json","view_paper":"https://pith.science/paper/NY2AY2XC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.15637&json=true","fetch_graph":"https://pith.science/api/pith-number/NY2AY2XCSU2JC6IDDCQW5DALM5/graph.json","fetch_events":"https://pith.science/api/pith-number/NY2AY2XCSU2JC6IDDCQW5DALM5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NY2AY2XCSU2JC6IDDCQW5DALM5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NY2AY2XCSU2JC6IDDCQW5DALM5/action/storage_attestation","attest_author":"https://pith.science/pith/NY2AY2XCSU2JC6IDDCQW5DALM5/action/author_attestation","sign_citation":"https://pith.science/pith/NY2AY2XCSU2JC6IDDCQW5DALM5/action/citation_signature","submit_replication":"https://pith.science/pith/NY2AY2XCSU2JC6IDDCQW5DALM5/action/replication_record"}},"created_at":"2026-07-01T00:17:07.881868+00:00","updated_at":"2026-07-01T00:17:07.881868+00:00"}