{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:RKRKOZTRLASR7S3ADMYND5HIY7","short_pith_number":"pith:RKRKOZTR","canonical_record":{"source":{"id":"2606.05610","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-04T02:32:11Z","cross_cats_sorted":[],"title_canon_sha256":"eacc065fee3d13feb121000e46efbbfcf36e1fc63d2b3ef34db9f84107feb371","abstract_canon_sha256":"c368591cd5183b7beb0e8d933d9e537f8663fe355852a9deb53dd531c4ff8662"},"schema_version":"1.0"},"canonical_sha256":"8aa2a7667158251fcb601b30d1f4e8c7f62bc2e4e58fc60feb38706c748b4903","source":{"kind":"arxiv","id":"2606.05610","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.05610","created_at":"2026-06-05T01:14:56Z"},{"alias_kind":"arxiv_version","alias_value":"2606.05610v1","created_at":"2026-06-05T01:14:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.05610","created_at":"2026-06-05T01:14:56Z"},{"alias_kind":"pith_short_12","alias_value":"RKRKOZTRLASR","created_at":"2026-06-05T01:14:56Z"},{"alias_kind":"pith_short_16","alias_value":"RKRKOZTRLASR7S3A","created_at":"2026-06-05T01:14:56Z"},{"alias_kind":"pith_short_8","alias_value":"RKRKOZTR","created_at":"2026-06-05T01:14:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:RKRKOZTRLASR7S3ADMYND5HIY7","target":"record","payload":{"canonical_record":{"source":{"id":"2606.05610","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-04T02:32:11Z","cross_cats_sorted":[],"title_canon_sha256":"eacc065fee3d13feb121000e46efbbfcf36e1fc63d2b3ef34db9f84107feb371","abstract_canon_sha256":"c368591cd5183b7beb0e8d933d9e537f8663fe355852a9deb53dd531c4ff8662"},"schema_version":"1.0"},"canonical_sha256":"8aa2a7667158251fcb601b30d1f4e8c7f62bc2e4e58fc60feb38706c748b4903","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T01:14:56.341680Z","signature_b64":"QFIIP+3BlHvndrDrtTV+mUDxteu8312bopRZ2QdyKYoSlgSdDvj+ihhySIIuMtnSIgUmZu/0OuMsnGHNYxFfCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8aa2a7667158251fcb601b30d1f4e8c7f62bc2e4e58fc60feb38706c748b4903","last_reissued_at":"2026-06-05T01:14:56.341201Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T01:14:56.341201Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.05610","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":"2026-06-05T01:14:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QJLKYIz5UgPVQmeWXv9AhKaT+I8rCRO9uM2CCfmvIumaMNeupPor9PBhM4YeFWa+4NICN1mtbktau/UR8D5+Ag==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-02T18:02:30.416684Z"},"content_sha256":"371bbae53b13629983ad7a5a93e6aab8aea86f715ef5872b68a160bfbbef4dbc","schema_version":"1.0","event_id":"sha256:371bbae53b13629983ad7a5a93e6aab8aea86f715ef5872b68a160bfbbef4dbc"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:RKRKOZTRLASR7S3ADMYND5HIY7","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Predictable Scaling Laws of Optimal Hyperparameters for LLM Continued Pre-training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Juncheng Diao, Junlin Shang, Peiguang Li, Rongxiang Weng, Yongwei Zhou","submitted_at":"2026-06-04T02:32:11Z","abstract_excerpt":"The efficacy of continued pre-training for Large Language Models (LLMs) hinges upon hyperparameter configurations, such as learning rate and batch size. However, current practices often rely on heuristics or grid searches, leading to training instability and excessive costs. In this work, we first empirically discover that optimal hyperparameters follow stable and predictable scaling laws throughout the continued pre-training process. Leveraging these insights, we propose a novel framework to establish quantitative relationships between compute budget and optimal hyperparameters for a given ch"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.05610","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/2606.05610/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":"2026-06-05T01:14:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zttMSVSPBeDuk9irCHRBqpidIfgX2RBiuLL2pZ7ElCyYZrrpnN5gvyrqdfwlhlHrUoVEhiJzLg7+LRILkUFNCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-02T18:02:30.417070Z"},"content_sha256":"74b668a14f98b88f7f54842c0665c2accc627d2089c48c4622497e57c142850e","schema_version":"1.0","event_id":"sha256:74b668a14f98b88f7f54842c0665c2accc627d2089c48c4622497e57c142850e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RKRKOZTRLASR7S3ADMYND5HIY7/bundle.json","state_url":"https://pith.science/pith/RKRKOZTRLASR7S3ADMYND5HIY7/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RKRKOZTRLASR7S3ADMYND5HIY7/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":"2026-07-02T18:02:30Z","links":{"resolver":"https://pith.science/pith/RKRKOZTRLASR7S3ADMYND5HIY7","bundle":"https://pith.science/pith/RKRKOZTRLASR7S3ADMYND5HIY7/bundle.json","state":"https://pith.science/pith/RKRKOZTRLASR7S3ADMYND5HIY7/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RKRKOZTRLASR7S3ADMYND5HIY7/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:RKRKOZTRLASR7S3ADMYND5HIY7","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":"c368591cd5183b7beb0e8d933d9e537f8663fe355852a9deb53dd531c4ff8662","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-04T02:32:11Z","title_canon_sha256":"eacc065fee3d13feb121000e46efbbfcf36e1fc63d2b3ef34db9f84107feb371"},"schema_version":"1.0","source":{"id":"2606.05610","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.05610","created_at":"2026-06-05T01:14:56Z"},{"alias_kind":"arxiv_version","alias_value":"2606.05610v1","created_at":"2026-06-05T01:14:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.05610","created_at":"2026-06-05T01:14:56Z"},{"alias_kind":"pith_short_12","alias_value":"RKRKOZTRLASR","created_at":"2026-06-05T01:14:56Z"},{"alias_kind":"pith_short_16","alias_value":"RKRKOZTRLASR7S3A","created_at":"2026-06-05T01:14:56Z"},{"alias_kind":"pith_short_8","alias_value":"RKRKOZTR","created_at":"2026-06-05T01:14:56Z"}],"graph_snapshots":[{"event_id":"sha256:74b668a14f98b88f7f54842c0665c2accc627d2089c48c4622497e57c142850e","target":"graph","created_at":"2026-06-05T01:14:56Z","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/2606.05610/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The efficacy of continued pre-training for Large Language Models (LLMs) hinges upon hyperparameter configurations, such as learning rate and batch size. However, current practices often rely on heuristics or grid searches, leading to training instability and excessive costs. In this work, we first empirically discover that optimal hyperparameters follow stable and predictable scaling laws throughout the continued pre-training process. Leveraging these insights, we propose a novel framework to establish quantitative relationships between compute budget and optimal hyperparameters for a given ch","authors_text":"Juncheng Diao, Junlin Shang, Peiguang Li, Rongxiang Weng, Yongwei Zhou","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-04T02:32:11Z","title":"Predictable Scaling Laws of Optimal Hyperparameters for LLM Continued Pre-training"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.05610","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":"sha256:371bbae53b13629983ad7a5a93e6aab8aea86f715ef5872b68a160bfbbef4dbc","target":"record","created_at":"2026-06-05T01:14:56Z","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":"c368591cd5183b7beb0e8d933d9e537f8663fe355852a9deb53dd531c4ff8662","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-04T02:32:11Z","title_canon_sha256":"eacc065fee3d13feb121000e46efbbfcf36e1fc63d2b3ef34db9f84107feb371"},"schema_version":"1.0","source":{"id":"2606.05610","kind":"arxiv","version":1}},"canonical_sha256":"8aa2a7667158251fcb601b30d1f4e8c7f62bc2e4e58fc60feb38706c748b4903","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8aa2a7667158251fcb601b30d1f4e8c7f62bc2e4e58fc60feb38706c748b4903","first_computed_at":"2026-06-05T01:14:56.341201Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-05T01:14:56.341201Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"QFIIP+3BlHvndrDrtTV+mUDxteu8312bopRZ2QdyKYoSlgSdDvj+ihhySIIuMtnSIgUmZu/0OuMsnGHNYxFfCA==","signature_status":"signed_v1","signed_at":"2026-06-05T01:14:56.341680Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.05610","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:371bbae53b13629983ad7a5a93e6aab8aea86f715ef5872b68a160bfbbef4dbc","sha256:74b668a14f98b88f7f54842c0665c2accc627d2089c48c4622497e57c142850e"],"state_sha256":"2825b066e40a2fef08423f3fb88e37530d843eab2abcda8e5802fd4b3b0e2ea6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rqFI2wGtqI8MYzduZKG9Nch97HtChYan2AVmRFsuY95DYP0ibhNo9HT+k2b9673oOgLuhavTKDKcfLpbf8dhDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-02T18:02:30.419270Z","bundle_sha256":"6efa3771a79b088c0e6c267d2da0f8f7b848586494da8034c6c8c26be047b15e"}}