{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:YTWPZ257LB4E7EOSYW535KG2HV","short_pith_number":"pith:YTWPZ257","schema_version":"1.0","canonical_sha256":"c4ecfcebbf58784f91d2c5bbbea8da3d6b2db6655d14bccf39b1e449a39b4a64","source":{"kind":"arxiv","id":"2409.13768","version":1},"attestation_state":"computed","paper":{"title":"Magika: AI-Powered Content-Type Detection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Alex Petit-Bianco, Ange Albertini, David Tao, Elie Bursztein, Francois Galilee, Giancarlo Metitieri, Julien Cretin, Kurt Thomas, Loua Farah, Luca Invernizzi, Marina Zhang, Yanick Fratantonio","submitted_at":"2024-09-18T17:24:39Z","abstract_excerpt":"The task of content-type detection -- which entails identifying the data encoded in an arbitrary byte sequence -- is critical for operating systems, development, reverse engineering environments, and a variety of security applications. In this paper, we introduce Magika, a novel AI-powered content-type detection tool. Under the hood, Magika employs a deep learning model that can execute on a single CPU with just 1MB of memory to store the model's weights. We show that Magika achieves an average F1 score of 99% across over a hundred content types and a test set of more than 1M files, outperform"},"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":"2409.13768","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2024-09-18T17:24:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"722e43364a3929784dcf4ed587e18946931323685b26b60b4e2823117ece7c56","abstract_canon_sha256":"52b54e02ec4f5dae5e364351d88d570872fbc4ce4fc62d45b65bbf8685ba8bee"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:10:06.561885Z","signature_b64":"zj5WELWJKGpX6SzX9OW0Lif8Pa+SW6N8d1FakPN9eVLQG8vkicL06uf0BgwIBE+Kj9a4SptbVxgGKbRBsH6wDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c4ecfcebbf58784f91d2c5bbbea8da3d6b2db6655d14bccf39b1e449a39b4a64","last_reissued_at":"2026-07-05T09:10:06.561404Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:10:06.561404Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Magika: AI-Powered Content-Type Detection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Alex Petit-Bianco, Ange Albertini, David Tao, Elie Bursztein, Francois Galilee, Giancarlo Metitieri, Julien Cretin, Kurt Thomas, Loua Farah, Luca Invernizzi, Marina Zhang, Yanick Fratantonio","submitted_at":"2024-09-18T17:24:39Z","abstract_excerpt":"The task of content-type detection -- which entails identifying the data encoded in an arbitrary byte sequence -- is critical for operating systems, development, reverse engineering environments, and a variety of security applications. In this paper, we introduce Magika, a novel AI-powered content-type detection tool. Under the hood, Magika employs a deep learning model that can execute on a single CPU with just 1MB of memory to store the model's weights. We show that Magika achieves an average F1 score of 99% across over a hundred content types and a test set of more than 1M files, outperform"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.13768","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/2409.13768/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":"2409.13768","created_at":"2026-07-05T09:10:06.561464+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.13768v1","created_at":"2026-07-05T09:10:06.561464+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.13768","created_at":"2026-07-05T09:10:06.561464+00:00"},{"alias_kind":"pith_short_12","alias_value":"YTWPZ257LB4E","created_at":"2026-07-05T09:10:06.561464+00:00"},{"alias_kind":"pith_short_16","alias_value":"YTWPZ257LB4E7EOS","created_at":"2026-07-05T09:10:06.561464+00:00"},{"alias_kind":"pith_short_8","alias_value":"YTWPZ257","created_at":"2026-07-05T09:10:06.561464+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.04171","citing_title":"MimeLens: Position-Agnostic Content-Type Detection for Binary Fragments","ref_index":4,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YTWPZ257LB4E7EOSYW535KG2HV","json":"https://pith.science/pith/YTWPZ257LB4E7EOSYW535KG2HV.json","graph_json":"https://pith.science/api/pith-number/YTWPZ257LB4E7EOSYW535KG2HV/graph.json","events_json":"https://pith.science/api/pith-number/YTWPZ257LB4E7EOSYW535KG2HV/events.json","paper":"https://pith.science/paper/YTWPZ257"},"agent_actions":{"view_html":"https://pith.science/pith/YTWPZ257LB4E7EOSYW535KG2HV","download_json":"https://pith.science/pith/YTWPZ257LB4E7EOSYW535KG2HV.json","view_paper":"https://pith.science/paper/YTWPZ257","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.13768&json=true","fetch_graph":"https://pith.science/api/pith-number/YTWPZ257LB4E7EOSYW535KG2HV/graph.json","fetch_events":"https://pith.science/api/pith-number/YTWPZ257LB4E7EOSYW535KG2HV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YTWPZ257LB4E7EOSYW535KG2HV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YTWPZ257LB4E7EOSYW535KG2HV/action/storage_attestation","attest_author":"https://pith.science/pith/YTWPZ257LB4E7EOSYW535KG2HV/action/author_attestation","sign_citation":"https://pith.science/pith/YTWPZ257LB4E7EOSYW535KG2HV/action/citation_signature","submit_replication":"https://pith.science/pith/YTWPZ257LB4E7EOSYW535KG2HV/action/replication_record"}},"created_at":"2026-07-05T09:10:06.561464+00:00","updated_at":"2026-07-05T09:10:06.561464+00:00"}