{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:37G2J3II3D2P2EPFJ2GB33D2VA","short_pith_number":"pith:37G2J3II","canonical_record":{"source":{"id":"2605.27354","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-26T17:55:59Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"947a3b576177f2dff21551a7b800a0cb3a62d015c8f580d43a9ba91882d77ef2","abstract_canon_sha256":"c0062ddcc8ff7469a8a8a21afd1b3dfeac32da237d98bccd88c0db0893db6a16"},"schema_version":"1.0"},"canonical_sha256":"dfcda4ed08d8f4fd11e54e8c1dec7aa838b1c38911859c62d34234b58cda1740","source":{"kind":"arxiv","id":"2605.27354","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.27354","created_at":"2026-05-27T02:06:19Z"},{"alias_kind":"arxiv_version","alias_value":"2605.27354v1","created_at":"2026-05-27T02:06:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.27354","created_at":"2026-05-27T02:06:19Z"},{"alias_kind":"pith_short_12","alias_value":"37G2J3II3D2P","created_at":"2026-05-27T02:06:19Z"},{"alias_kind":"pith_short_16","alias_value":"37G2J3II3D2P2EPF","created_at":"2026-05-27T02:06:19Z"},{"alias_kind":"pith_short_8","alias_value":"37G2J3II","created_at":"2026-05-27T02:06:19Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:37G2J3II3D2P2EPFJ2GB33D2VA","target":"record","payload":{"canonical_record":{"source":{"id":"2605.27354","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-26T17:55:59Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"947a3b576177f2dff21551a7b800a0cb3a62d015c8f580d43a9ba91882d77ef2","abstract_canon_sha256":"c0062ddcc8ff7469a8a8a21afd1b3dfeac32da237d98bccd88c0db0893db6a16"},"schema_version":"1.0"},"canonical_sha256":"dfcda4ed08d8f4fd11e54e8c1dec7aa838b1c38911859c62d34234b58cda1740","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-27T02:06:19.458919Z","signature_b64":"8X8RCkS/XDZM6ixtsBbL4SK6/X+3l7gqGsW/CHCOqnIZXLdqKhvIaWsulxkc2Sc3nF7TzhsFvn9vjMXycv9lDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dfcda4ed08d8f4fd11e54e8c1dec7aa838b1c38911859c62d34234b58cda1740","last_reissued_at":"2026-05-27T02:06:19.458167Z","signature_status":"signed_v1","first_computed_at":"2026-05-27T02:06:19.458167Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.27354","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":"2026-05-27T02:06:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JhhlbQZmhrrxXrsC+YpgYKdSwGAFwVoJ8pBWHAdwa8j6XjXLJK35yfH3IdFaWOcnqUwesGtn6sS/G/Njy3ooAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T08:57:30.051618Z"},"content_sha256":"e8e55837d308c425429c1db5879d936e60ca8d05bff07f979131671cb1178ceb","schema_version":"1.0","event_id":"sha256:e8e55837d308c425429c1db5879d936e60ca8d05bff07f979131671cb1178ceb"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:37G2J3II3D2P2EPFJ2GB33D2VA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Guiding LLM Post-training Data Engineering with Model Internals from Sparse Autoencoders","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Jinwu Hu, Juanzi Li, Lei Hou, Xiaozhi Wang, Yi Jing, Zao Dai, Zijun Yao","submitted_at":"2026-05-26T17:55:59Z","abstract_excerpt":"Model internals encode rich information about how a large language model (LLM) processes its training data; however, post-training data engineering largely relies on external signals and ignores rich intrinsic signals lying in model internals. We propose SAERL, a data engineering framework for LLM reinforcement learning (RL). It models three intrinsic data properties: diversity, difficulty, and quality, using model internals extracted with Sparse Autoencoder (SAE), an advanced mechanistic interpretability tool. Each property grounds a concrete data engineering operation: SAE-space clustering w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.27354","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/2605.27354/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":"2026-05-27T02:06:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mcHo9pPQKDFwwBddHm6bF/tHX7mfisPLHOqoAkdw6wl/J6zE8/uZwpad6565RyCODkzq+coPtAO0GlJ5E4CIDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T08:57:30.052270Z"},"content_sha256":"400e93d789972858f80b13a1a7d891a3ac21c93dddb4769ddd3c8924a66490ff","schema_version":"1.0","event_id":"sha256:400e93d789972858f80b13a1a7d891a3ac21c93dddb4769ddd3c8924a66490ff"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/37G2J3II3D2P2EPFJ2GB33D2VA/bundle.json","state_url":"https://pith.science/pith/37G2J3II3D2P2EPFJ2GB33D2VA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/37G2J3II3D2P2EPFJ2GB33D2VA/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":"2026-05-30T08:57:30Z","links":{"resolver":"https://pith.science/pith/37G2J3II3D2P2EPFJ2GB33D2VA","bundle":"https://pith.science/pith/37G2J3II3D2P2EPFJ2GB33D2VA/bundle.json","state":"https://pith.science/pith/37G2J3II3D2P2EPFJ2GB33D2VA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/37G2J3II3D2P2EPFJ2GB33D2VA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:37G2J3II3D2P2EPFJ2GB33D2VA","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":"c0062ddcc8ff7469a8a8a21afd1b3dfeac32da237d98bccd88c0db0893db6a16","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-26T17:55:59Z","title_canon_sha256":"947a3b576177f2dff21551a7b800a0cb3a62d015c8f580d43a9ba91882d77ef2"},"schema_version":"1.0","source":{"id":"2605.27354","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.27354","created_at":"2026-05-27T02:06:19Z"},{"alias_kind":"arxiv_version","alias_value":"2605.27354v1","created_at":"2026-05-27T02:06:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.27354","created_at":"2026-05-27T02:06:19Z"},{"alias_kind":"pith_short_12","alias_value":"37G2J3II3D2P","created_at":"2026-05-27T02:06:19Z"},{"alias_kind":"pith_short_16","alias_value":"37G2J3II3D2P2EPF","created_at":"2026-05-27T02:06:19Z"},{"alias_kind":"pith_short_8","alias_value":"37G2J3II","created_at":"2026-05-27T02:06:19Z"}],"graph_snapshots":[{"event_id":"sha256:400e93d789972858f80b13a1a7d891a3ac21c93dddb4769ddd3c8924a66490ff","target":"graph","created_at":"2026-05-27T02:06:19Z","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/2605.27354/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Model internals encode rich information about how a large language model (LLM) processes its training data; however, post-training data engineering largely relies on external signals and ignores rich intrinsic signals lying in model internals. We propose SAERL, a data engineering framework for LLM reinforcement learning (RL). It models three intrinsic data properties: diversity, difficulty, and quality, using model internals extracted with Sparse Autoencoder (SAE), an advanced mechanistic interpretability tool. Each property grounds a concrete data engineering operation: SAE-space clustering w","authors_text":"Jinwu Hu, Juanzi Li, Lei Hou, Xiaozhi Wang, Yi Jing, Zao Dai, Zijun Yao","cross_cats":["cs.AI","cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-26T17:55:59Z","title":"Guiding LLM Post-training Data Engineering with Model Internals from Sparse Autoencoders"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.27354","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:e8e55837d308c425429c1db5879d936e60ca8d05bff07f979131671cb1178ceb","target":"record","created_at":"2026-05-27T02:06:19Z","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":"c0062ddcc8ff7469a8a8a21afd1b3dfeac32da237d98bccd88c0db0893db6a16","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-26T17:55:59Z","title_canon_sha256":"947a3b576177f2dff21551a7b800a0cb3a62d015c8f580d43a9ba91882d77ef2"},"schema_version":"1.0","source":{"id":"2605.27354","kind":"arxiv","version":1}},"canonical_sha256":"dfcda4ed08d8f4fd11e54e8c1dec7aa838b1c38911859c62d34234b58cda1740","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dfcda4ed08d8f4fd11e54e8c1dec7aa838b1c38911859c62d34234b58cda1740","first_computed_at":"2026-05-27T02:06:19.458167Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-27T02:06:19.458167Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8X8RCkS/XDZM6ixtsBbL4SK6/X+3l7gqGsW/CHCOqnIZXLdqKhvIaWsulxkc2Sc3nF7TzhsFvn9vjMXycv9lDA==","signature_status":"signed_v1","signed_at":"2026-05-27T02:06:19.458919Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.27354","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e8e55837d308c425429c1db5879d936e60ca8d05bff07f979131671cb1178ceb","sha256:400e93d789972858f80b13a1a7d891a3ac21c93dddb4769ddd3c8924a66490ff"],"state_sha256":"0224b901f392780f303d74c04a910ecaa7c2f47820fe1497733592cbdeb5afaf"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FfF4H6NMukRUpehzzwCOWa6jwyWo/E8F2uJ7dWV68Gwnk0ZkIWB3xxBI4gvvbRiyrSbVm0GJp5SqRrMm6oSuAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-30T08:57:30.054697Z","bundle_sha256":"2e3286683cd06327cab3355d76f5146ccc266487769cc5b13a19a110cf53fa37"}}