{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:HLMWWSUVFKVRYTBC3AUNU4W3HS","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":"d0c90fbe97ac1abe83c2f4b2fb4db97256e9c09f754e46150368d48a92fc55c3","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-25T07:15:38Z","title_canon_sha256":"5987de6c94874e9012523988b38cadb30c3999c9cf5faa96ee7f19095396781e"},"schema_version":"1.0","source":{"id":"2605.25511","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.25511","created_at":"2026-05-26T02:04:40Z"},{"alias_kind":"arxiv_version","alias_value":"2605.25511v1","created_at":"2026-05-26T02:04:40Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.25511","created_at":"2026-05-26T02:04:40Z"},{"alias_kind":"pith_short_12","alias_value":"HLMWWSUVFKVR","created_at":"2026-05-26T02:04:40Z"},{"alias_kind":"pith_short_16","alias_value":"HLMWWSUVFKVRYTBC","created_at":"2026-05-26T02:04:40Z"},{"alias_kind":"pith_short_8","alias_value":"HLMWWSUV","created_at":"2026-05-26T02:04:40Z"}],"graph_snapshots":[{"event_id":"sha256:8b815463117d39b73a0f17efed16d877663a5751c686854b62c25a4d694fd5b1","target":"graph","created_at":"2026-05-26T02:04:40Z","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/2605.25511/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent advancements in Reinforcement Learning (RL), particularly Group Relative Policy Optimization (GRPO), have significantly enhanced the reasoning capabilities of Large Language Models. However, applying these problem-centric optimization methods to role-playing agents often leads to a loss of character fidelity and style collapse, as they prioritize context-specific utility over persona alignment. To address this, we propose Character-Centric Group Relative Policy Optimization (CRPO), a framework designed to realign RL objectives with the role-playing task. CRPO improves character distinct","authors_text":"Benyou Wang, Kehai Chen, Liang Yue, Min Zhang, Yihong Tang","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-25T07:15:38Z","title":"CRPO: Character-centric Group Relative Policy Optimization for Role-aware Reasoning in Role-playing Agents"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.25511","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":"sha256:76cda533faa958a3868de2ad3b4cd36345a1dee994c6efa43a483fc67b3eea71","target":"record","created_at":"2026-05-26T02:04:40Z","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":"d0c90fbe97ac1abe83c2f4b2fb4db97256e9c09f754e46150368d48a92fc55c3","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-25T07:15:38Z","title_canon_sha256":"5987de6c94874e9012523988b38cadb30c3999c9cf5faa96ee7f19095396781e"},"schema_version":"1.0","source":{"id":"2605.25511","kind":"arxiv","version":1}},"canonical_sha256":"3ad96b4a952aab1c4c22d828da72db3c90dbb9a541be885f8b02219d240fff4c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3ad96b4a952aab1c4c22d828da72db3c90dbb9a541be885f8b02219d240fff4c","first_computed_at":"2026-05-26T02:04:40.217195Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-26T02:04:40.217195Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"mBxtgAe5ipay0v6Qm1mREcbCdnYL6xPee5ys7kecYz0JoxO7uajFU7vQkggWM0cQmMg10sUqta0fx1HHzbxzBQ==","signature_status":"signed_v1","signed_at":"2026-05-26T02:04:40.218123Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.25511","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:76cda533faa958a3868de2ad3b4cd36345a1dee994c6efa43a483fc67b3eea71","sha256:8b815463117d39b73a0f17efed16d877663a5751c686854b62c25a4d694fd5b1"],"state_sha256":"4da505a6b43bd2622a5ebc0e919b5f6bd5467dfa22a0dae3c41c9fea13faca48"}