{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:GAZURJBZ5A3HMTP4Y5MRM23EGS","short_pith_number":"pith:GAZURJBZ","schema_version":"1.0","canonical_sha256":"303348a439e836764dfcc759166b6434bf5bbd846d416b4ac733cd5ec9073ccf","source":{"kind":"arxiv","id":"2606.01126","version":1},"attestation_state":"computed","paper":{"title":"STARFISH: faST Accuracy Recovery in pruned networks From Internal State Healing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Adi Shamir, Odelia Melamed, Shir Maon","submitted_at":"2026-05-31T09:59:08Z","abstract_excerpt":"Pruning is a process designed to reduce the number of weights in a large neural network. This can substantially speed up inference but might cause a considerable reduction in the model's accuracy, and thus it is usually followed by a healing process that regains some of the lost accuracy. In this paper, we propose a new healing method, STARFISH, that can recover (most of) the accuracy of any pruned network efficiently. The main idea of STARFISH is to optimize the pruned network to align with the original network's internal state representations using a tiny calibration set of unlabeled example"},"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":"2606.01126","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-31T09:59:08Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"8221e9bdae96c87ea7d61de79752bc13258d7abf902c0c2b88df8203c7b0ec50","abstract_canon_sha256":"25d74e5f0c33e27ee3ced15d207c50fdc76b5ee93b0d2637e51db818b2544425"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:24.245924Z","signature_b64":"5RxzCpe7W+1lk2hc19NEYpIcJ2iQjp1ag/dELEkNOcvKhjgp/0PF+jo2q3x0o83ZHLqW5xRjWmJm7sA0HDfWDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"303348a439e836764dfcc759166b6434bf5bbd846d416b4ac733cd5ec9073ccf","last_reissued_at":"2026-06-02T02:04:24.245410Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:24.245410Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"STARFISH: faST Accuracy Recovery in pruned networks From Internal State Healing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Adi Shamir, Odelia Melamed, Shir Maon","submitted_at":"2026-05-31T09:59:08Z","abstract_excerpt":"Pruning is a process designed to reduce the number of weights in a large neural network. This can substantially speed up inference but might cause a considerable reduction in the model's accuracy, and thus it is usually followed by a healing process that regains some of the lost accuracy. In this paper, we propose a new healing method, STARFISH, that can recover (most of) the accuracy of any pruned network efficiently. The main idea of STARFISH is to optimize the pruned network to align with the original network's internal state representations using a tiny calibration set of unlabeled example"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.01126","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/2606.01126/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":"2606.01126","created_at":"2026-06-02T02:04:24.245473+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.01126v1","created_at":"2026-06-02T02:04:24.245473+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.01126","created_at":"2026-06-02T02:04:24.245473+00:00"},{"alias_kind":"pith_short_12","alias_value":"GAZURJBZ5A3H","created_at":"2026-06-02T02:04:24.245473+00:00"},{"alias_kind":"pith_short_16","alias_value":"GAZURJBZ5A3HMTP4","created_at":"2026-06-02T02:04:24.245473+00:00"},{"alias_kind":"pith_short_8","alias_value":"GAZURJBZ","created_at":"2026-06-02T02:04:24.245473+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GAZURJBZ5A3HMTP4Y5MRM23EGS","json":"https://pith.science/pith/GAZURJBZ5A3HMTP4Y5MRM23EGS.json","graph_json":"https://pith.science/api/pith-number/GAZURJBZ5A3HMTP4Y5MRM23EGS/graph.json","events_json":"https://pith.science/api/pith-number/GAZURJBZ5A3HMTP4Y5MRM23EGS/events.json","paper":"https://pith.science/paper/GAZURJBZ"},"agent_actions":{"view_html":"https://pith.science/pith/GAZURJBZ5A3HMTP4Y5MRM23EGS","download_json":"https://pith.science/pith/GAZURJBZ5A3HMTP4Y5MRM23EGS.json","view_paper":"https://pith.science/paper/GAZURJBZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.01126&json=true","fetch_graph":"https://pith.science/api/pith-number/GAZURJBZ5A3HMTP4Y5MRM23EGS/graph.json","fetch_events":"https://pith.science/api/pith-number/GAZURJBZ5A3HMTP4Y5MRM23EGS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GAZURJBZ5A3HMTP4Y5MRM23EGS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GAZURJBZ5A3HMTP4Y5MRM23EGS/action/storage_attestation","attest_author":"https://pith.science/pith/GAZURJBZ5A3HMTP4Y5MRM23EGS/action/author_attestation","sign_citation":"https://pith.science/pith/GAZURJBZ5A3HMTP4Y5MRM23EGS/action/citation_signature","submit_replication":"https://pith.science/pith/GAZURJBZ5A3HMTP4Y5MRM23EGS/action/replication_record"}},"created_at":"2026-06-02T02:04:24.245473+00:00","updated_at":"2026-06-02T02:04:24.245473+00:00"}