{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:4553VIKCWOU6GQ3KUDPULHFIQG","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":"ee2dc46a48737d881c77c25ea2cbf18b3f9523b86d8bdb13173120911df8ae92","cross_cats_sorted":["cs.CR","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-07-20T14:47:03Z","title_canon_sha256":"0b33d13080add23998e147b2632c03e68b6573e12593eae70536630ae0c8ab5f"},"schema_version":"1.0","source":{"id":"2507.14987","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2507.14987","created_at":"2026-07-05T11:40:16Z"},{"alias_kind":"arxiv_version","alias_value":"2507.14987v1","created_at":"2026-07-05T11:40:16Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.14987","created_at":"2026-07-05T11:40:16Z"},{"alias_kind":"pith_short_12","alias_value":"4553VIKCWOU6","created_at":"2026-07-05T11:40:16Z"},{"alias_kind":"pith_short_16","alias_value":"4553VIKCWOU6GQ3K","created_at":"2026-07-05T11:40:16Z"},{"alias_kind":"pith_short_8","alias_value":"4553VIKC","created_at":"2026-07-05T11:40:16Z"}],"graph_snapshots":[{"event_id":"sha256:b37583acb125bf44694f06bd52833b057d8ca8acfe06186b6bf1f032c53ee232","target":"graph","created_at":"2026-07-05T11:40:16Z","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/2507.14987/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs), despite possessing latent safety understanding from their vast pretraining data, remain vulnerable to generating harmful content and exhibit issues such as over-refusal and utility degradation after safety alignment. Current safety alignment methods often result in superficial refusal shortcuts or rely on intensive supervision for reasoning-based approaches, failing to fully leverage the model's intrinsic safety self-awareness. We propose \\textbf{AlphaAlign}, a simple yet effective pure reinforcement learning (RL) framework with verifiable safety reward designed t","authors_text":"An Zhang, Leheng Sheng, Xiang Wang, XiuYu Zhang, Yi Zhang, Yuxin Chen, Zhenkai Liang","cross_cats":["cs.CR","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-07-20T14:47:03Z","title":"AlphaAlign: Incentivizing Safety Alignment with Extremely Simplified Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.14987","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":"sha256:41a3aeed7a467b017bc24b1e02879d1f7d0e9d277f170119dda981594f15a2f4","target":"record","created_at":"2026-07-05T11:40:16Z","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":"ee2dc46a48737d881c77c25ea2cbf18b3f9523b86d8bdb13173120911df8ae92","cross_cats_sorted":["cs.CR","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-07-20T14:47:03Z","title_canon_sha256":"0b33d13080add23998e147b2632c03e68b6573e12593eae70536630ae0c8ab5f"},"schema_version":"1.0","source":{"id":"2507.14987","kind":"arxiv","version":1}},"canonical_sha256":"e77bbaa142b3a9e3436aa0df459ca881ae636e0d0b211bbfcc01e073832f63bf","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e77bbaa142b3a9e3436aa0df459ca881ae636e0d0b211bbfcc01e073832f63bf","first_computed_at":"2026-07-05T11:40:16.284280Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:40:16.284280Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"RRgb37Dv9JrsKXuJ4Ksda9Bt+bcAH+yOrjOZWAD0/mWHpAuQfYljgP1TWG6WQYYE4SGx6NRQxPy+ephKsOK5BQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:40:16.284776Z","signed_message":"canonical_sha256_bytes"},"source_id":"2507.14987","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:41a3aeed7a467b017bc24b1e02879d1f7d0e9d277f170119dda981594f15a2f4","sha256:b37583acb125bf44694f06bd52833b057d8ca8acfe06186b6bf1f032c53ee232"],"state_sha256":"d6d5d4e130ac8cc2054135b0debb0056599822b7a47fd849b4aadc2010f6870d"}