{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:7L2QCRDTCF5GS6TCIJRGNUVMC4","short_pith_number":"pith:7L2QCRDT","schema_version":"1.0","canonical_sha256":"faf5014473117a697a62426266d2ac170d124034d68b797c6f1d8375f7bfeed7","source":{"kind":"arxiv","id":"2407.13493","version":4},"attestation_state":"computed","paper":{"title":"Training Foundation Models as Data Compression: On Information, Model Weights and Copyright Law","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CY","authors_text":"Claudia Cevenini, Giorgio Franceschelli, Mirco Musolesi","submitted_at":"2024-07-18T13:23:16Z","abstract_excerpt":"The training process of foundation models as for other classes of deep learning systems is based on minimizing the reconstruction error over a training set. For this reason, they are susceptible to the memorization and subsequent reproduction of training samples. In this paper, we introduce a training-as-compressing perspective, wherein the model's weights embody a compressed representation of the training data. From a copyright standpoint, this point of view implies that the weights can be considered a reproduction or, more likely, a derivative work of a potentially protected set of works. We"},"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":"2407.13493","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CY","submitted_at":"2024-07-18T13:23:16Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"92140d34541f9744399eba99e28e3cedd398e00e4924eda7b65f0ccbe528cc82","abstract_canon_sha256":"d3688dee145af825199d4890d3b9c5a0f5110c13f60bb26147f989faf2f3e616"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:29:19.742217Z","signature_b64":"et0F/hHOCxBt/ggBks5g/W42DHvlZeNM/lPrJQCJhDBIzOJSoEFRrPqB2Nj2OuJxkqU0KMygveGy/a3KI8sHCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"faf5014473117a697a62426266d2ac170d124034d68b797c6f1d8375f7bfeed7","last_reissued_at":"2026-07-05T10:29:19.741272Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:29:19.741272Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training Foundation Models as Data Compression: On Information, Model Weights and Copyright Law","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CY","authors_text":"Claudia Cevenini, Giorgio Franceschelli, Mirco Musolesi","submitted_at":"2024-07-18T13:23:16Z","abstract_excerpt":"The training process of foundation models as for other classes of deep learning systems is based on minimizing the reconstruction error over a training set. For this reason, they are susceptible to the memorization and subsequent reproduction of training samples. In this paper, we introduce a training-as-compressing perspective, wherein the model's weights embody a compressed representation of the training data. From a copyright standpoint, this point of view implies that the weights can be considered a reproduction or, more likely, a derivative work of a potentially protected set of works. We"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.13493","kind":"arxiv","version":4},"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/2407.13493/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":"2407.13493","created_at":"2026-07-05T10:29:19.741399+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.13493v4","created_at":"2026-07-05T10:29:19.741399+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.13493","created_at":"2026-07-05T10:29:19.741399+00:00"},{"alias_kind":"pith_short_12","alias_value":"7L2QCRDTCF5G","created_at":"2026-07-05T10:29:19.741399+00:00"},{"alias_kind":"pith_short_16","alias_value":"7L2QCRDTCF5GS6TC","created_at":"2026-07-05T10:29:19.741399+00:00"},{"alias_kind":"pith_short_8","alias_value":"7L2QCRDT","created_at":"2026-07-05T10:29:19.741399+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.03547","citing_title":"Erase Persona, Forget Lore: Benchmarking Multimodal Copyright Unlearning in Large Vision Language Models","ref_index":32,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7L2QCRDTCF5GS6TCIJRGNUVMC4","json":"https://pith.science/pith/7L2QCRDTCF5GS6TCIJRGNUVMC4.json","graph_json":"https://pith.science/api/pith-number/7L2QCRDTCF5GS6TCIJRGNUVMC4/graph.json","events_json":"https://pith.science/api/pith-number/7L2QCRDTCF5GS6TCIJRGNUVMC4/events.json","paper":"https://pith.science/paper/7L2QCRDT"},"agent_actions":{"view_html":"https://pith.science/pith/7L2QCRDTCF5GS6TCIJRGNUVMC4","download_json":"https://pith.science/pith/7L2QCRDTCF5GS6TCIJRGNUVMC4.json","view_paper":"https://pith.science/paper/7L2QCRDT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.13493&json=true","fetch_graph":"https://pith.science/api/pith-number/7L2QCRDTCF5GS6TCIJRGNUVMC4/graph.json","fetch_events":"https://pith.science/api/pith-number/7L2QCRDTCF5GS6TCIJRGNUVMC4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7L2QCRDTCF5GS6TCIJRGNUVMC4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7L2QCRDTCF5GS6TCIJRGNUVMC4/action/storage_attestation","attest_author":"https://pith.science/pith/7L2QCRDTCF5GS6TCIJRGNUVMC4/action/author_attestation","sign_citation":"https://pith.science/pith/7L2QCRDTCF5GS6TCIJRGNUVMC4/action/citation_signature","submit_replication":"https://pith.science/pith/7L2QCRDTCF5GS6TCIJRGNUVMC4/action/replication_record"}},"created_at":"2026-07-05T10:29:19.741399+00:00","updated_at":"2026-07-05T10:29:19.741399+00:00"}