{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:IJOWF3PN35QZQEB3XGOMC5RRSZ","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"9e9b38ead1492a4a78e134842478b7fe77435bb1f41e2f2f0645c729b9d148f9","cross_cats_sorted":["cs.AI","cs.CL","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-12T17:52:47Z","title_canon_sha256":"5393fac0308bbeae126c14fb570fe3594a1e991399162925b1a18051af4ce5ec"},"schema_version":"1.0","source":{"id":"2502.08606","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.08606","created_at":"2026-07-05T11:43:09Z"},{"alias_kind":"arxiv_version","alias_value":"2502.08606v2","created_at":"2026-07-05T11:43:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.08606","created_at":"2026-07-05T11:43:09Z"},{"alias_kind":"pith_short_12","alias_value":"IJOWF3PN35QZ","created_at":"2026-07-05T11:43:09Z"},{"alias_kind":"pith_short_16","alias_value":"IJOWF3PN35QZQEB3","created_at":"2026-07-05T11:43:09Z"},{"alias_kind":"pith_short_8","alias_value":"IJOWF3PN","created_at":"2026-07-05T11:43:09Z"}],"graph_snapshots":[{"event_id":"sha256:3e2360fb208936ba397f6dcdf0c44987adb0d9dbb8aeee9fafe88b7392419782","target":"graph","created_at":"2026-07-05T11:43:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2502.08606/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We propose a distillation scaling law that estimates distilled model performance based on a compute budget and its allocation between the student and teacher. Our findings mitigate the risks associated with large-scale distillation by enabling compute-optimal allocation for both the teacher and student to maximize student performance. We provide compute-optimal distillation recipes for two key scenarios: when a teacher already exists, and when a teacher needs training. In settings involving many students or an existing teacher, distillation outperforms supervised learning up to a compute level","authors_text":"Amitis Shidani, Dan Busbridge, Etai Littwin, Floris Weers, Jason Ramapuram, Russ Webb","cross_cats":["cs.AI","cs.CL","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-12T17:52:47Z","title":"Distillation Scaling Laws"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.08606","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:5aafbad8e8e5cc30bc3fad1533baa28000c26a49ba93b0e22df5d8ae4062dce6","target":"record","created_at":"2026-07-05T11:43:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"9e9b38ead1492a4a78e134842478b7fe77435bb1f41e2f2f0645c729b9d148f9","cross_cats_sorted":["cs.AI","cs.CL","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-02-12T17:52:47Z","title_canon_sha256":"5393fac0308bbeae126c14fb570fe3594a1e991399162925b1a18051af4ce5ec"},"schema_version":"1.0","source":{"id":"2502.08606","kind":"arxiv","version":2}},"canonical_sha256":"425d62ededdf6198103bb99cc17631967f71e9fbc3aef0eb288566619e775480","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"425d62ededdf6198103bb99cc17631967f71e9fbc3aef0eb288566619e775480","first_computed_at":"2026-07-05T11:43:09.658382Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:43:09.658382Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"e1ZRo1XlxvVYPkDHFE6dXjj8GUHbQK+Pq+blEdFjzQJL/WOtSEXEgwzngL9nPQR6u8TFVJ/oIhb4bZCVheX6CQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:43:09.658912Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.08606","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5aafbad8e8e5cc30bc3fad1533baa28000c26a49ba93b0e22df5d8ae4062dce6","sha256:3e2360fb208936ba397f6dcdf0c44987adb0d9dbb8aeee9fafe88b7392419782"],"state_sha256":"76ec296ab4a6f948d39a215875a7a039dd9a7e7ac053818ecd93f5917d970485"}