{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:Z7XOP3B47XPWQPI7H3OVTZL2VC","short_pith_number":"pith:Z7XOP3B4","canonical_record":{"source":{"id":"2502.01187","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-02-03T09:23:53Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"5d04d40b1ae33d530c792d17d80f03e741f23a7f745969fd26cca5fe3764580f","abstract_canon_sha256":"c4e104e5459f7f1d6b57e77262e34c14a6a3bdb8a277d76d6a0621da2645c861"},"schema_version":"1.0"},"canonical_sha256":"cfeee7ec3cfddf683d1f3edd59e57aa8b6b881f48a8184b68e6d03420044af34","source":{"kind":"arxiv","id":"2502.01187","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.01187","created_at":"2026-07-05T10:08:43Z"},{"alias_kind":"arxiv_version","alias_value":"2502.01187v1","created_at":"2026-07-05T10:08:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.01187","created_at":"2026-07-05T10:08:43Z"},{"alias_kind":"pith_short_12","alias_value":"Z7XOP3B47XPW","created_at":"2026-07-05T10:08:43Z"},{"alias_kind":"pith_short_16","alias_value":"Z7XOP3B47XPWQPI7","created_at":"2026-07-05T10:08:43Z"},{"alias_kind":"pith_short_8","alias_value":"Z7XOP3B4","created_at":"2026-07-05T10:08:43Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:Z7XOP3B47XPWQPI7H3OVTZL2VC","target":"record","payload":{"canonical_record":{"source":{"id":"2502.01187","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-02-03T09:23:53Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"5d04d40b1ae33d530c792d17d80f03e741f23a7f745969fd26cca5fe3764580f","abstract_canon_sha256":"c4e104e5459f7f1d6b57e77262e34c14a6a3bdb8a277d76d6a0621da2645c861"},"schema_version":"1.0"},"canonical_sha256":"cfeee7ec3cfddf683d1f3edd59e57aa8b6b881f48a8184b68e6d03420044af34","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:08:43.935415Z","signature_b64":"dw5qW9YBQL1T8tN5n2wA12Xw4MM1AZYd0ppo63QcRSyHjZSqiORc3XzlADGP7B2x8LMkO7TpmDzVqXe1FpI+AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cfeee7ec3cfddf683d1f3edd59e57aa8b6b881f48a8184b68e6d03420044af34","last_reissued_at":"2026-07-05T10:08:43.935002Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:08:43.935002Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.01187","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":"2026-07-05T10:08:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"WRliAXnl8U6LUcYTTzKp8VH6NvGYw2ha5jK6Uj3+b3FKXu1gtKsV8g80huoHNd+62RE7PkzQ+P49BIrwvOX1CA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T10:59:13.455116Z"},"content_sha256":"51f899695bb968537178b5daf66c7554ab0292883b9914cc8d63c402283e5e9a","schema_version":"1.0","event_id":"sha256:51f899695bb968537178b5daf66c7554ab0292883b9914cc8d63c402283e5e9a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:Z7XOP3B47XPWQPI7H3OVTZL2VC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Skewed Memorization in Large Language Models: Quantification and Decomposition","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Amir M. Rahmani, Di Huang, Hao Li, Ziyu Wang","submitted_at":"2025-02-03T09:23:53Z","abstract_excerpt":"Memorization in Large Language Models (LLMs) poses privacy and security risks, as models may unintentionally reproduce sensitive or copyrighted data. Existing analyses focus on average-case scenarios, often neglecting the highly skewed distribution of memorization. This paper examines memorization in LLM supervised fine-tuning (SFT), exploring its relationships with training duration, dataset size, and inter-sample similarity. By analyzing memorization probabilities over sequence lengths, we link this skewness to the token generation process, offering insights for estimating memorization and c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.01187","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/2502.01187/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":"2026-07-05T10:08:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lYpGCprxIdajOdDMY7wgmPMC17MC4PGPzHyEs9TNEWjYxRVjR/Lqbs/pShnnH8BJ1aPsmYHi92mxxy0ccuJSBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T10:59:13.455670Z"},"content_sha256":"11719df9c78c7ed64f35d9105a25d5436e3ceea29ce81204c75217ac93733a51","schema_version":"1.0","event_id":"sha256:11719df9c78c7ed64f35d9105a25d5436e3ceea29ce81204c75217ac93733a51"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/Z7XOP3B47XPWQPI7H3OVTZL2VC/bundle.json","state_url":"https://pith.science/pith/Z7XOP3B47XPWQPI7H3OVTZL2VC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/Z7XOP3B47XPWQPI7H3OVTZL2VC/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":"2026-08-10T10:59:13Z","links":{"resolver":"https://pith.science/pith/Z7XOP3B47XPWQPI7H3OVTZL2VC","bundle":"https://pith.science/pith/Z7XOP3B47XPWQPI7H3OVTZL2VC/bundle.json","state":"https://pith.science/pith/Z7XOP3B47XPWQPI7H3OVTZL2VC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/Z7XOP3B47XPWQPI7H3OVTZL2VC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:Z7XOP3B47XPWQPI7H3OVTZL2VC","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":"c4e104e5459f7f1d6b57e77262e34c14a6a3bdb8a277d76d6a0621da2645c861","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-02-03T09:23:53Z","title_canon_sha256":"5d04d40b1ae33d530c792d17d80f03e741f23a7f745969fd26cca5fe3764580f"},"schema_version":"1.0","source":{"id":"2502.01187","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.01187","created_at":"2026-07-05T10:08:43Z"},{"alias_kind":"arxiv_version","alias_value":"2502.01187v1","created_at":"2026-07-05T10:08:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.01187","created_at":"2026-07-05T10:08:43Z"},{"alias_kind":"pith_short_12","alias_value":"Z7XOP3B47XPW","created_at":"2026-07-05T10:08:43Z"},{"alias_kind":"pith_short_16","alias_value":"Z7XOP3B47XPWQPI7","created_at":"2026-07-05T10:08:43Z"},{"alias_kind":"pith_short_8","alias_value":"Z7XOP3B4","created_at":"2026-07-05T10:08:43Z"}],"graph_snapshots":[{"event_id":"sha256:11719df9c78c7ed64f35d9105a25d5436e3ceea29ce81204c75217ac93733a51","target":"graph","created_at":"2026-07-05T10:08:43Z","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/2502.01187/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Memorization in Large Language Models (LLMs) poses privacy and security risks, as models may unintentionally reproduce sensitive or copyrighted data. Existing analyses focus on average-case scenarios, often neglecting the highly skewed distribution of memorization. This paper examines memorization in LLM supervised fine-tuning (SFT), exploring its relationships with training duration, dataset size, and inter-sample similarity. By analyzing memorization probabilities over sequence lengths, we link this skewness to the token generation process, offering insights for estimating memorization and c","authors_text":"Amir M. Rahmani, Di Huang, Hao Li, Ziyu Wang","cross_cats":["cs.CL","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-02-03T09:23:53Z","title":"Skewed Memorization in Large Language Models: Quantification and Decomposition"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.01187","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":"sha256:51f899695bb968537178b5daf66c7554ab0292883b9914cc8d63c402283e5e9a","target":"record","created_at":"2026-07-05T10:08:43Z","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":"c4e104e5459f7f1d6b57e77262e34c14a6a3bdb8a277d76d6a0621da2645c861","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-02-03T09:23:53Z","title_canon_sha256":"5d04d40b1ae33d530c792d17d80f03e741f23a7f745969fd26cca5fe3764580f"},"schema_version":"1.0","source":{"id":"2502.01187","kind":"arxiv","version":1}},"canonical_sha256":"cfeee7ec3cfddf683d1f3edd59e57aa8b6b881f48a8184b68e6d03420044af34","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"cfeee7ec3cfddf683d1f3edd59e57aa8b6b881f48a8184b68e6d03420044af34","first_computed_at":"2026-07-05T10:08:43.935002Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:08:43.935002Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"dw5qW9YBQL1T8tN5n2wA12Xw4MM1AZYd0ppo63QcRSyHjZSqiORc3XzlADGP7B2x8LMkO7TpmDzVqXe1FpI+AQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:08:43.935415Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.01187","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:51f899695bb968537178b5daf66c7554ab0292883b9914cc8d63c402283e5e9a","sha256:11719df9c78c7ed64f35d9105a25d5436e3ceea29ce81204c75217ac93733a51"],"state_sha256":"3f502e9390e266cd313fb64fa28691d818d6f7c3df47820e1295ad3664345571"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Lg8FweamPbXRVI9lXhP/YuUoBEBrU9cJpTa/T6zqTjeWK4yjLb5NTHyRUHsP+F6KhZNCMsdwjIrInMvbl2LrDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T10:59:13.459765Z","bundle_sha256":"3328071d3bfce4282f667a559dd0ed761a0b02b6bf6df91cd4f35d6ef041cd3a"}}