{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:ORG2DZ5OLIJ5KGWEJEOSQ6SFX4","short_pith_number":"pith:ORG2DZ5O","canonical_record":{"source":{"id":"2605.20105","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-19T16:56:56Z","cross_cats_sorted":[],"title_canon_sha256":"5c8f6335628d93922b09fc08ebe5b9b2a12386efe4e629fb7b889a4aaf54d6d6","abstract_canon_sha256":"c4e5a1838346ad3f0ba00cacfb5065786814ca395d78fdeb64b5e077c57160f8"},"schema_version":"1.0"},"canonical_sha256":"744da1e7ae5a13d51ac4491d287a45bf0950dadfca8e1d130c00345f2ec3d4c7","source":{"kind":"arxiv","id":"2605.20105","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.20105","created_at":"2026-05-20T02:06:02Z"},{"alias_kind":"arxiv_version","alias_value":"2605.20105v1","created_at":"2026-05-20T02:06:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.20105","created_at":"2026-05-20T02:06:02Z"},{"alias_kind":"pith_short_12","alias_value":"ORG2DZ5OLIJ5","created_at":"2026-05-20T02:06:02Z"},{"alias_kind":"pith_short_16","alias_value":"ORG2DZ5OLIJ5KGWE","created_at":"2026-05-20T02:06:02Z"},{"alias_kind":"pith_short_8","alias_value":"ORG2DZ5O","created_at":"2026-05-20T02:06:02Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:ORG2DZ5OLIJ5KGWEJEOSQ6SFX4","target":"record","payload":{"canonical_record":{"source":{"id":"2605.20105","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-19T16:56:56Z","cross_cats_sorted":[],"title_canon_sha256":"5c8f6335628d93922b09fc08ebe5b9b2a12386efe4e629fb7b889a4aaf54d6d6","abstract_canon_sha256":"c4e5a1838346ad3f0ba00cacfb5065786814ca395d78fdeb64b5e077c57160f8"},"schema_version":"1.0"},"canonical_sha256":"744da1e7ae5a13d51ac4491d287a45bf0950dadfca8e1d130c00345f2ec3d4c7","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T02:06:02.264043Z","signature_b64":"FmG7SpvI+1VVXVC6eBjlar29S9ZL6Zx+UeWBrB1HzjLhTK84KxOZUNjqJVLRUOAe4LlrbVhQH5uukHyu36Q1CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"744da1e7ae5a13d51ac4491d287a45bf0950dadfca8e1d130c00345f2ec3d4c7","last_reissued_at":"2026-05-20T02:06:02.263416Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T02:06:02.263416Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.20105","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T02:06:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4HaeRpcKeXOqxgZ3pBRD6/WhiPQ5FaK0wrv+vfH8CJLlZZRKHY3o+9+KmgPBaMMXdSJ1jkO9UErrJ+mcldKHCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-24T18:15:05.940229Z"},"content_sha256":"289b668bcf4ec4cb763f82961c98bbe212aab912d7d149f3d095d091a08a8850","schema_version":"1.0","event_id":"sha256:289b668bcf4ec4cb763f82961c98bbe212aab912d7d149f3d095d091a08a8850"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:ORG2DZ5OLIJ5KGWEJEOSQ6SFX4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Optimal Representation Size: High-Dimensional Analysis of Pretraining and Linear Probing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Andrew Saxe, Cl\\'ementine Domin\\'e, Marco Mondelli, Rachel Swanson, Valentina Njaradi","submitted_at":"2026-05-19T16:56:56Z","abstract_excerpt":"Learning to generalise from limited data is a fundamental challenge for both artificial and biological systems. A common strategy is to extract reusable structure from abundant unlabelled data, enabling efficient adaptation to new tasks from limited labelled data. This two-stage paradigm is now standard in modern training pipelines, where pretraining is followed by fine-tuning or linear probing. We provide an analytical model of this process: structure extraction is formalized as principal component analysis on unlabelled data, and downstream learning as linear regression on a separate labelle"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.20105","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/2605.20105/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T02:06:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PXiojoD2P6nD5+HhX42Qb+n5X1IQkUkmui2rO+GwhMH8Jg8KYLN5ZpqJNQQJ/YnG0mjd65tFSqv6mb6Q2T0ZCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-24T18:15:05.940941Z"},"content_sha256":"7a45f854d49270987a97bbf05f47bf92ec766a6dcf43ffa1d12b3bfaf4dd4fba","schema_version":"1.0","event_id":"sha256:7a45f854d49270987a97bbf05f47bf92ec766a6dcf43ffa1d12b3bfaf4dd4fba"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ORG2DZ5OLIJ5KGWEJEOSQ6SFX4/bundle.json","state_url":"https://pith.science/pith/ORG2DZ5OLIJ5KGWEJEOSQ6SFX4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ORG2DZ5OLIJ5KGWEJEOSQ6SFX4/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-24T18:15:05Z","links":{"resolver":"https://pith.science/pith/ORG2DZ5OLIJ5KGWEJEOSQ6SFX4","bundle":"https://pith.science/pith/ORG2DZ5OLIJ5KGWEJEOSQ6SFX4/bundle.json","state":"https://pith.science/pith/ORG2DZ5OLIJ5KGWEJEOSQ6SFX4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ORG2DZ5OLIJ5KGWEJEOSQ6SFX4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:ORG2DZ5OLIJ5KGWEJEOSQ6SFX4","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":"c4e5a1838346ad3f0ba00cacfb5065786814ca395d78fdeb64b5e077c57160f8","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-19T16:56:56Z","title_canon_sha256":"5c8f6335628d93922b09fc08ebe5b9b2a12386efe4e629fb7b889a4aaf54d6d6"},"schema_version":"1.0","source":{"id":"2605.20105","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.20105","created_at":"2026-05-20T02:06:02Z"},{"alias_kind":"arxiv_version","alias_value":"2605.20105v1","created_at":"2026-05-20T02:06:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.20105","created_at":"2026-05-20T02:06:02Z"},{"alias_kind":"pith_short_12","alias_value":"ORG2DZ5OLIJ5","created_at":"2026-05-20T02:06:02Z"},{"alias_kind":"pith_short_16","alias_value":"ORG2DZ5OLIJ5KGWE","created_at":"2026-05-20T02:06:02Z"},{"alias_kind":"pith_short_8","alias_value":"ORG2DZ5O","created_at":"2026-05-20T02:06:02Z"}],"graph_snapshots":[{"event_id":"sha256:7a45f854d49270987a97bbf05f47bf92ec766a6dcf43ffa1d12b3bfaf4dd4fba","target":"graph","created_at":"2026-05-20T02:06:02Z","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/2605.20105/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Learning to generalise from limited data is a fundamental challenge for both artificial and biological systems. A common strategy is to extract reusable structure from abundant unlabelled data, enabling efficient adaptation to new tasks from limited labelled data. This two-stage paradigm is now standard in modern training pipelines, where pretraining is followed by fine-tuning or linear probing. We provide an analytical model of this process: structure extraction is formalized as principal component analysis on unlabelled data, and downstream learning as linear regression on a separate labelle","authors_text":"Andrew Saxe, Cl\\'ementine Domin\\'e, Marco Mondelli, Rachel Swanson, Valentina Njaradi","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-19T16:56:56Z","title":"Optimal Representation Size: High-Dimensional Analysis of Pretraining and Linear Probing"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.20105","kind":"arxiv","version":1},"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":"sha256:289b668bcf4ec4cb763f82961c98bbe212aab912d7d149f3d095d091a08a8850","target":"record","created_at":"2026-05-20T02:06:02Z","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":"c4e5a1838346ad3f0ba00cacfb5065786814ca395d78fdeb64b5e077c57160f8","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-19T16:56:56Z","title_canon_sha256":"5c8f6335628d93922b09fc08ebe5b9b2a12386efe4e629fb7b889a4aaf54d6d6"},"schema_version":"1.0","source":{"id":"2605.20105","kind":"arxiv","version":1}},"canonical_sha256":"744da1e7ae5a13d51ac4491d287a45bf0950dadfca8e1d130c00345f2ec3d4c7","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"744da1e7ae5a13d51ac4491d287a45bf0950dadfca8e1d130c00345f2ec3d4c7","first_computed_at":"2026-05-20T02:06:02.263416Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T02:06:02.263416Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"FmG7SpvI+1VVXVC6eBjlar29S9ZL6Zx+UeWBrB1HzjLhTK84KxOZUNjqJVLRUOAe4LlrbVhQH5uukHyu36Q1CA==","signature_status":"signed_v1","signed_at":"2026-05-20T02:06:02.264043Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.20105","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:289b668bcf4ec4cb763f82961c98bbe212aab912d7d149f3d095d091a08a8850","sha256:7a45f854d49270987a97bbf05f47bf92ec766a6dcf43ffa1d12b3bfaf4dd4fba"],"state_sha256":"2e4c4b03a422cc8bd6fd2d1434244fbda2cb5a9db8d399d0ea7c1f56a4b69f85"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"CASdWvuM6OAwaQCDKv3K23YfBARoD8QYH5YDSYZaL9gaMj8Mkdq4xHknJLPFTDqGjsqFFfj37UwJ7Cl2Cg1aDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-24T18:15:05.944716Z","bundle_sha256":"6ea405e1269345bfdef39751b642249713a2368066dd59da1df63172ef111e24"}}