{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:6WXOD5SM2SNANSAT65L4TCTSBR","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":"1f4248b069c48c97e934a6043f2e762d904813be4c5e8a72120ee598992d82cb","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-30T14:38:49Z","title_canon_sha256":"86757466a917e368817b15ab1a3fc43054a6403469998b863922083c24ce871c"},"schema_version":"1.0","source":{"id":"2606.31748","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.31748","created_at":"2026-07-01T01:18:13Z"},{"alias_kind":"arxiv_version","alias_value":"2606.31748v1","created_at":"2026-07-01T01:18:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.31748","created_at":"2026-07-01T01:18:13Z"},{"alias_kind":"pith_short_12","alias_value":"6WXOD5SM2SNA","created_at":"2026-07-01T01:18:13Z"},{"alias_kind":"pith_short_16","alias_value":"6WXOD5SM2SNANSAT","created_at":"2026-07-01T01:18:13Z"},{"alias_kind":"pith_short_8","alias_value":"6WXOD5SM","created_at":"2026-07-01T01:18:13Z"}],"graph_snapshots":[{"event_id":"sha256:98a4b570053a5c0591ed82a1fb9ae0ff2a90be5b0d4b0b059731d2ecc46a8329","target":"graph","created_at":"2026-07-01T01:18:13Z","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/2606.31748/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Safety training on language models often induces over-refusal: improved safety on harmful prompts at the cost of increased refusal on harmless ones. Though this trade-off can be mitigated by training models with reinforcement learning (RL) to reason before answering, it does not remove the underlying problem that reasoning can often be a \"rubber stamp\" for a predetermined response. In this paper, we address the safety-refusal trade-off by rethinking how models are trained to reason about safety. Our key insight is that unsafe reasoning can itself serve as a useful exploratory signal. Rather th","authors_text":"Aviral Kumar, Taeyoun Kim","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-30T14:38:49Z","title":"Addressing Over-Refusal in LLMs with Competing Rewards"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.31748","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:c498da7b22f8f7cafb3d3666ea6b1947b585d9c0e5d13da6a001e03e0e6a1f13","target":"record","created_at":"2026-07-01T01:18:13Z","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":"1f4248b069c48c97e934a6043f2e762d904813be4c5e8a72120ee598992d82cb","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-30T14:38:49Z","title_canon_sha256":"86757466a917e368817b15ab1a3fc43054a6403469998b863922083c24ce871c"},"schema_version":"1.0","source":{"id":"2606.31748","kind":"arxiv","version":1}},"canonical_sha256":"f5aee1f64cd49a06c813f757c98a720c56fb6f3542e6e3d6590e535d28d087ee","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f5aee1f64cd49a06c813f757c98a720c56fb6f3542e6e3d6590e535d28d087ee","first_computed_at":"2026-07-01T01:18:13.620090Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-01T01:18:13.620090Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"TbVxeO2D9DSMd3gx4GQ5lD+qOK3zIawyw2/oapDIFNfUKIfKZ397OyxTv+p1TouMpjolJbR0gyDG4wFA1ZKpDw==","signature_status":"signed_v1","signed_at":"2026-07-01T01:18:13.620512Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.31748","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c498da7b22f8f7cafb3d3666ea6b1947b585d9c0e5d13da6a001e03e0e6a1f13","sha256:98a4b570053a5c0591ed82a1fb9ae0ff2a90be5b0d4b0b059731d2ecc46a8329"],"state_sha256":"e2d4182a8c7d557c7d4a331fb0611a09715862aa2f7e5ff4e57c94956fc1b2dd"}