{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:WMFHIL7VUGJN2Z3RXLIJ2IM4CY","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":"ce9dbd7c0ef454f4ae641de317082b3fc1ac5ad4efe77a412909cc1923cdeea7","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T05:40:22Z","title_canon_sha256":"a82fe9fc24db1cbb34a2c61975eb8ead7ddd0f7d8daf9c72239fd80d723d2f02"},"schema_version":"1.0","source":{"id":"2605.27996","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.27996","created_at":"2026-05-28T01:04:55Z"},{"alias_kind":"arxiv_version","alias_value":"2605.27996v1","created_at":"2026-05-28T01:04:55Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.27996","created_at":"2026-05-28T01:04:55Z"},{"alias_kind":"pith_short_12","alias_value":"WMFHIL7VUGJN","created_at":"2026-05-28T01:04:55Z"},{"alias_kind":"pith_short_16","alias_value":"WMFHIL7VUGJN2Z3R","created_at":"2026-05-28T01:04:55Z"},{"alias_kind":"pith_short_8","alias_value":"WMFHIL7V","created_at":"2026-05-28T01:04:55Z"}],"graph_snapshots":[{"event_id":"sha256:c94f7e4905f35c9c1a201c532d2bbd5eafadc9ffb6e39b3a7ee1efae9fa49e73","target":"graph","created_at":"2026-05-28T01:04:55Z","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/2605.27996/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Single-axis mitigations of reward-model biases (e.g., reducing proxy reliance on length, sycophancy, or style) can rotate optimization pressure onto correlated proxies rather than eliminate it, a failure mode we call reward bias substitution. The failure is enabled by a measurement-versus-optimization gap between audit and policy-induced distributions during mitigation evaluation and policy training. We formalize mitigation outcomes into a regime taxonomy and prove that successful mitigation, bias substitution, and overcorrection produce identical observables under any audit-distribution scori","authors_text":"Andreas Haupt, Daniel Fein, Marcel Hussing, Max Lamparth, Mykel J. Kochenderfer","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T05:40:22Z","title":"Reward Bias Substitution: Single-Axis Bias Mitigations Redirect Optimization Pressure"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.27996","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:d1d6ec0f439726c2add37ff140dd39e7e6bc086993368dfce81ce32d7bccc4dd","target":"record","created_at":"2026-05-28T01:04:55Z","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":"ce9dbd7c0ef454f4ae641de317082b3fc1ac5ad4efe77a412909cc1923cdeea7","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-27T05:40:22Z","title_canon_sha256":"a82fe9fc24db1cbb34a2c61975eb8ead7ddd0f7d8daf9c72239fd80d723d2f02"},"schema_version":"1.0","source":{"id":"2605.27996","kind":"arxiv","version":1}},"canonical_sha256":"b30a742ff5a192dd6771bad09d219c1609e8a359721c6efc80ac66bb41b933de","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b30a742ff5a192dd6771bad09d219c1609e8a359721c6efc80ac66bb41b933de","first_computed_at":"2026-05-28T01:04:55.578254Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-28T01:04:55.578254Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"C/baupogX9GWnzehl23eTHfsWN5kP6tGA4IeqRyHdL4EFNpyDfvAfuCrnL68EFZy5tyBoGmRFm1VKzzcY9eKBQ==","signature_status":"signed_v1","signed_at":"2026-05-28T01:04:55.578727Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.27996","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d1d6ec0f439726c2add37ff140dd39e7e6bc086993368dfce81ce32d7bccc4dd","sha256:c94f7e4905f35c9c1a201c532d2bbd5eafadc9ffb6e39b3a7ee1efae9fa49e73"],"state_sha256":"eae2f42162a781359d752ef738f3c6de3b2cda54ecdb53f372baa920cf175f9e"}