{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:SP22TY2PMTFJMW42EBIYT3XPPT","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":"286cda8fdc35c9d38dcc2e37f7736b18a12a27f1379897f5e02b953a9dc5cf22","cross_cats_sorted":["cs.LG","stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-09T17:23:29Z","title_canon_sha256":"a675701def5eaaab2ae50b2e3bb3cde31d3bfb23e5902aa8947758c51630e321"},"schema_version":"1.0","source":{"id":"2409.05816","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.05816","created_at":"2026-07-05T10:28:05Z"},{"alias_kind":"arxiv_version","alias_value":"2409.05816v2","created_at":"2026-07-05T10:28:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.05816","created_at":"2026-07-05T10:28:05Z"},{"alias_kind":"pith_short_12","alias_value":"SP22TY2PMTFJ","created_at":"2026-07-05T10:28:05Z"},{"alias_kind":"pith_short_16","alias_value":"SP22TY2PMTFJMW42","created_at":"2026-07-05T10:28:05Z"},{"alias_kind":"pith_short_8","alias_value":"SP22TY2P","created_at":"2026-07-05T10:28:05Z"}],"graph_snapshots":[{"event_id":"sha256:36c282761dae9830643cc394641db8b43c44fff0a1a31a6fdf551a1eb686abe1","target":"graph","created_at":"2026-07-05T10:28:05Z","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/2409.05816/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Quality pretraining data is often seen as the key to high-performance language models. However, progress in understanding pretraining data has been slow due to the costly pretraining runs required for data selection experiments. We present a framework that avoids these costs and selects high-quality pretraining data without any LLM training of our own. Our work is based on a simple observation: LLM losses on many pretraining texts are correlated with downstream benchmark performance, and selecting high-correlation documents is an effective pretraining data selection method. We build a new stat","authors_text":"Christopher Potts, Tatsunori Hashimoto, Tristan Thrush","cross_cats":["cs.LG","stat.ML"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-09T17:23:29Z","title":"Improving Pretraining Data Using Perplexity Correlations"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.05816","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":"sha256:328aa07408de30a51d4d738a85b81fe24ec857bee68fdc48ce4edf79c5b62274","target":"record","created_at":"2026-07-05T10:28:05Z","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":"286cda8fdc35c9d38dcc2e37f7736b18a12a27f1379897f5e02b953a9dc5cf22","cross_cats_sorted":["cs.LG","stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-09T17:23:29Z","title_canon_sha256":"a675701def5eaaab2ae50b2e3bb3cde31d3bfb23e5902aa8947758c51630e321"},"schema_version":"1.0","source":{"id":"2409.05816","kind":"arxiv","version":2}},"canonical_sha256":"93f5a9e34f64ca965b9a205189eeef7cc5f170135845c3700170f5e593295d81","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"93f5a9e34f64ca965b9a205189eeef7cc5f170135845c3700170f5e593295d81","first_computed_at":"2026-07-05T10:28:05.017891Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:28:05.017891Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hpHC9IwnoaQpJmuycVGhhNh4OUeTSq29ZxMD7n+13dsBnGvdKKdg2FnWn46aEZbHyfjF1MavMfbYlacDlCG1Ag==","signature_status":"signed_v1","signed_at":"2026-07-05T10:28:05.018858Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.05816","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:328aa07408de30a51d4d738a85b81fe24ec857bee68fdc48ce4edf79c5b62274","sha256:36c282761dae9830643cc394641db8b43c44fff0a1a31a6fdf551a1eb686abe1"],"state_sha256":"bd267d704e373c4006ec2db97c077f67251fff93ad6e0cc5f082a7b996480410"}