{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:3TNI45ZJJBU63RBYTI2FLPWUYY","short_pith_number":"pith:3TNI45ZJ","canonical_record":{"source":{"id":"2505.16260","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-22T05:50:19Z","cross_cats_sorted":[],"title_canon_sha256":"513d15ead7be70b3bce1621b3976558806babde39c75c9d165a1cc6a0a91e262","abstract_canon_sha256":"e2bc41ae09223cc236ed3abf9a05adc8c2877ebcf112e6bd6c266a159415d2ab"},"schema_version":"1.0"},"canonical_sha256":"dcda8e77294869edc4389a3455bed4c6183caafa484e216b9046509fe9629e51","source":{"kind":"arxiv","id":"2505.16260","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.16260","created_at":"2026-07-05T11:07:32Z"},{"alias_kind":"arxiv_version","alias_value":"2505.16260v1","created_at":"2026-07-05T11:07:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.16260","created_at":"2026-07-05T11:07:32Z"},{"alias_kind":"pith_short_12","alias_value":"3TNI45ZJJBU6","created_at":"2026-07-05T11:07:32Z"},{"alias_kind":"pith_short_16","alias_value":"3TNI45ZJJBU63RBY","created_at":"2026-07-05T11:07:32Z"},{"alias_kind":"pith_short_8","alias_value":"3TNI45ZJ","created_at":"2026-07-05T11:07:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:3TNI45ZJJBU63RBYTI2FLPWUYY","target":"record","payload":{"canonical_record":{"source":{"id":"2505.16260","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-22T05:50:19Z","cross_cats_sorted":[],"title_canon_sha256":"513d15ead7be70b3bce1621b3976558806babde39c75c9d165a1cc6a0a91e262","abstract_canon_sha256":"e2bc41ae09223cc236ed3abf9a05adc8c2877ebcf112e6bd6c266a159415d2ab"},"schema_version":"1.0"},"canonical_sha256":"dcda8e77294869edc4389a3455bed4c6183caafa484e216b9046509fe9629e51","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:07:32.324167Z","signature_b64":"xK7rH1zFbNIFo804Aioga+s/BsoILCh/ST0DBuZQeqoV2txwIQzKmPlWOZuuwzm+cXglWyR5dzXjD4C9Dz2GBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dcda8e77294869edc4389a3455bed4c6183caafa484e216b9046509fe9629e51","last_reissued_at":"2026-07-05T11:07:32.323786Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:07:32.323786Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.16260","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":"2026-07-05T11:07:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uABSylOfc2kXmTX5hExmgdCbxruTSoJTrFJ9QP4dabFyEFSbIN3VrN8yAAM6zbznydr7CfwvZUcHJCaQrXaTCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T21:01:59.324503Z"},"content_sha256":"e079cbb18a4f34e0d63e83031d9724fbf87479132c29f27527d8c05afdb78fb2","schema_version":"1.0","event_id":"sha256:e079cbb18a4f34e0d63e83031d9724fbf87479132c29f27527d8c05afdb78fb2"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:3TNI45ZJJBU63RBYTI2FLPWUYY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Small-to-Large Generalization: Data Influences Models Consistently Across Scale","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Alaa Khaddaj, Aleksander Madry, Logan Engstrom","submitted_at":"2025-05-22T05:50:19Z","abstract_excerpt":"Choice of training data distribution greatly influences model behavior. Yet, in large-scale settings, precisely characterizing how changes in training data affects predictions is often difficult due to model training costs. Current practice is to instead extrapolate from scaled down, inexpensive-to-train proxy models. However, changes in data do not influence smaller and larger models identically. Therefore, understanding how choice of data affects large-scale models raises the question: how does training data distribution influence model behavior across compute scale? We find that small- and "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.16260","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/2505.16260/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":"2026-07-05T11:07:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YaUiHtPRfrOSI6DbnSdY7iwqQuxcUv2FlpUnJ884Ez+hnRx7bMxQ34uMlgs8IsQ45TV4Y7T3Dy7uvLRDzTMmAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T21:01:59.325373Z"},"content_sha256":"902a81695f3a51bfc59b57b6cc9fd5b2d0c4e7bed6d56bd2ace7e6419e8bf6c3","schema_version":"1.0","event_id":"sha256:902a81695f3a51bfc59b57b6cc9fd5b2d0c4e7bed6d56bd2ace7e6419e8bf6c3"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3TNI45ZJJBU63RBYTI2FLPWUYY/bundle.json","state_url":"https://pith.science/pith/3TNI45ZJJBU63RBYTI2FLPWUYY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3TNI45ZJJBU63RBYTI2FLPWUYY/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":"2026-08-07T21:01:59Z","links":{"resolver":"https://pith.science/pith/3TNI45ZJJBU63RBYTI2FLPWUYY","bundle":"https://pith.science/pith/3TNI45ZJJBU63RBYTI2FLPWUYY/bundle.json","state":"https://pith.science/pith/3TNI45ZJJBU63RBYTI2FLPWUYY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3TNI45ZJJBU63RBYTI2FLPWUYY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:3TNI45ZJJBU63RBYTI2FLPWUYY","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":"e2bc41ae09223cc236ed3abf9a05adc8c2877ebcf112e6bd6c266a159415d2ab","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-22T05:50:19Z","title_canon_sha256":"513d15ead7be70b3bce1621b3976558806babde39c75c9d165a1cc6a0a91e262"},"schema_version":"1.0","source":{"id":"2505.16260","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.16260","created_at":"2026-07-05T11:07:32Z"},{"alias_kind":"arxiv_version","alias_value":"2505.16260v1","created_at":"2026-07-05T11:07:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.16260","created_at":"2026-07-05T11:07:32Z"},{"alias_kind":"pith_short_12","alias_value":"3TNI45ZJJBU6","created_at":"2026-07-05T11:07:32Z"},{"alias_kind":"pith_short_16","alias_value":"3TNI45ZJJBU63RBY","created_at":"2026-07-05T11:07:32Z"},{"alias_kind":"pith_short_8","alias_value":"3TNI45ZJ","created_at":"2026-07-05T11:07:32Z"}],"graph_snapshots":[{"event_id":"sha256:902a81695f3a51bfc59b57b6cc9fd5b2d0c4e7bed6d56bd2ace7e6419e8bf6c3","target":"graph","created_at":"2026-07-05T11:07:32Z","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/2505.16260/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Choice of training data distribution greatly influences model behavior. Yet, in large-scale settings, precisely characterizing how changes in training data affects predictions is often difficult due to model training costs. Current practice is to instead extrapolate from scaled down, inexpensive-to-train proxy models. However, changes in data do not influence smaller and larger models identically. Therefore, understanding how choice of data affects large-scale models raises the question: how does training data distribution influence model behavior across compute scale? We find that small- and ","authors_text":"Alaa Khaddaj, Aleksander Madry, Logan Engstrom","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-22T05:50:19Z","title":"Small-to-Large Generalization: Data Influences Models Consistently Across Scale"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.16260","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:e079cbb18a4f34e0d63e83031d9724fbf87479132c29f27527d8c05afdb78fb2","target":"record","created_at":"2026-07-05T11:07:32Z","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":"e2bc41ae09223cc236ed3abf9a05adc8c2877ebcf112e6bd6c266a159415d2ab","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-22T05:50:19Z","title_canon_sha256":"513d15ead7be70b3bce1621b3976558806babde39c75c9d165a1cc6a0a91e262"},"schema_version":"1.0","source":{"id":"2505.16260","kind":"arxiv","version":1}},"canonical_sha256":"dcda8e77294869edc4389a3455bed4c6183caafa484e216b9046509fe9629e51","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dcda8e77294869edc4389a3455bed4c6183caafa484e216b9046509fe9629e51","first_computed_at":"2026-07-05T11:07:32.323786Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:07:32.323786Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"xK7rH1zFbNIFo804Aioga+s/BsoILCh/ST0DBuZQeqoV2txwIQzKmPlWOZuuwzm+cXglWyR5dzXjD4C9Dz2GBg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:07:32.324167Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.16260","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e079cbb18a4f34e0d63e83031d9724fbf87479132c29f27527d8c05afdb78fb2","sha256:902a81695f3a51bfc59b57b6cc9fd5b2d0c4e7bed6d56bd2ace7e6419e8bf6c3"],"state_sha256":"954833cc9dc2cf6a5e2e7e355e0a2f6d6d357b48e4bf06b2fd2d4b91225a372b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XzdiJdmEd6+fLMERoYNf1GIOHUOl630SJp8mjRi+UI34zumSQEqFXAXPKtF9NquVkTubcmIzq5yuf7YJ74kBCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T21:01:59.332365Z","bundle_sha256":"793d6d4d66b3712581e282cff0bb382c1ec849ba4b927bdcfc2dc3a6ed9c689f"}}