{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:NBUNZJQQ7FGR2LAHRK6PDLQAZP","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":"6dd5a269065e0265c66bd202739a804ada7ab3430ceb6762578a9a1cbba6274d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-09T14:44:01Z","title_canon_sha256":"e5c47ff66b21c965e23e6f69e3d097c12d9554471cb32d8fd8c0a4d224544d06"},"schema_version":"1.0","source":{"id":"2606.10931","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.10931","created_at":"2026-06-10T01:10:48Z"},{"alias_kind":"arxiv_version","alias_value":"2606.10931v1","created_at":"2026-06-10T01:10:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.10931","created_at":"2026-06-10T01:10:48Z"},{"alias_kind":"pith_short_12","alias_value":"NBUNZJQQ7FGR","created_at":"2026-06-10T01:10:48Z"},{"alias_kind":"pith_short_16","alias_value":"NBUNZJQQ7FGR2LAH","created_at":"2026-06-10T01:10:48Z"},{"alias_kind":"pith_short_8","alias_value":"NBUNZJQQ","created_at":"2026-06-10T01:10:48Z"}],"graph_snapshots":[{"event_id":"sha256:c5548e01d4dacfa9efada998d8442376e2467ead0e1db861ebb6e098884a3f3a","target":"graph","created_at":"2026-06-10T01:10:48Z","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/2606.10931/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Warning: This paper contains several toxic and offensive statements.\n  Modern large language models (LLMs) are typically aligned through large-scale post-training to ensure fair and reliable behavior. In this work, we investigate how easily such guardrails can be broken by Group Relative Policy Optimization (GRPO). We show that one-shot GRPO training on a single biased example is sufficient to induce systematic bias, with stereotype-driven reasoning generalizing across attributes, categories, and benchmarks. We further find that models differ in their susceptibility based on the initial likeli","authors_text":"Clayton Scott, Naichen Shi, Naihao Deng, Rada Mihalcea, Yilun Zhu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-09T14:44:01Z","title":"It Takes One to Bias Them All: Breaking Bad with One-Shot GRPO"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.10931","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":"sha256:5fb7ef2a8ee5b8741ece6c8c6fd2fb8244f779a93b175460dda10c0738bf44ef","target":"record","created_at":"2026-06-10T01:10:48Z","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":"6dd5a269065e0265c66bd202739a804ada7ab3430ceb6762578a9a1cbba6274d","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-09T14:44:01Z","title_canon_sha256":"e5c47ff66b21c965e23e6f69e3d097c12d9554471cb32d8fd8c0a4d224544d06"},"schema_version":"1.0","source":{"id":"2606.10931","kind":"arxiv","version":1}},"canonical_sha256":"6868dca610f94d1d2c078abcf1ae00cbc407be06da17ce2b1c315176583f2fe0","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6868dca610f94d1d2c078abcf1ae00cbc407be06da17ce2b1c315176583f2fe0","first_computed_at":"2026-06-10T01:10:48.720648Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-10T01:10:48.720648Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZqXof7deStScM30xv74z9JJiyzQPoKKyrUxgZqEqDWa6MaIk6HCs6FAQTe1A4RzxMP9ihNXe0romPQVS48xeAw==","signature_status":"signed_v1","signed_at":"2026-06-10T01:10:48.721503Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.10931","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5fb7ef2a8ee5b8741ece6c8c6fd2fb8244f779a93b175460dda10c0738bf44ef","sha256:c5548e01d4dacfa9efada998d8442376e2467ead0e1db861ebb6e098884a3f3a"],"state_sha256":"f7cca3d89e512836d7891195b7c4f61dc72fde8fe4a403ddd6f9dcc3b42c500f"}