{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:LQB5USZGAMNLLIZ6TUUUCV55NL","short_pith_number":"pith:LQB5USZG","schema_version":"1.0","canonical_sha256":"5c03da4b26031ab5a33e9d294157bd6af72ebd5b3cb70408090859568672a652","source":{"kind":"arxiv","id":"2605.29190","version":1},"attestation_state":"computed","paper":{"title":"When RL Suppresses Its Own Vocabulary: Recovering Reasoning Diversity in Puzzle-to-Math Transfer","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Arjun Karuvally, Fergal Reid, Mayug Maniparambil, Terrence Sejnowski","submitted_at":"2026-05-28T00:04:18Z","abstract_excerpt":"Reinforcement learning using verifiable rewards (RLVR) improves LLM reasoning, but the conditions under which it transfers across domains -- and why it does so -- remain under-explored. We study cross-domain transfer in a 7B model whose SFT and RL post-training stages use only constraint-satisfaction puzzles, with no mathematics problems in the post-training data. To analyze how transfer emerges, we introduce a reasoning primitive-level framework that combines a 9-class span classifier with motif extraction, allowing us to segment chain-of-thought traces into primitive motifs and track their e"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2605.29190","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-28T00:04:18Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"d4303d94efb9dad98d5eb4b9d69aba0fe1c619779f3c4c60eaa2b91ada3f02f1","abstract_canon_sha256":"1d6cd4576436f13d7ad4a968d42b1e895c5cc11f208c15f4016779af4b034e36"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T01:05:23.465614Z","signature_b64":"58fkMotJrR7SHEzWSF6qxYdBaMBCw3gsdoo0iTuNrlKBU2Tf6ZFtsvrkXC5Hiii0wbU8T1MU1ZCO6Dz6xTPEDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5c03da4b26031ab5a33e9d294157bd6af72ebd5b3cb70408090859568672a652","last_reissued_at":"2026-05-29T01:05:23.464756Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T01:05:23.464756Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"When RL Suppresses Its Own Vocabulary: Recovering Reasoning Diversity in Puzzle-to-Math Transfer","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Arjun Karuvally, Fergal Reid, Mayug Maniparambil, Terrence Sejnowski","submitted_at":"2026-05-28T00:04:18Z","abstract_excerpt":"Reinforcement learning using verifiable rewards (RLVR) improves LLM reasoning, but the conditions under which it transfers across domains -- and why it does so -- remain under-explored. We study cross-domain transfer in a 7B model whose SFT and RL post-training stages use only constraint-satisfaction puzzles, with no mathematics problems in the post-training data. To analyze how transfer emerges, we introduce a reasoning primitive-level framework that combines a 9-class span classifier with motif extraction, allowing us to segment chain-of-thought traces into primitive motifs and track their e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.29190","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.29190/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2605.29190","created_at":"2026-05-29T01:05:23.464890+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.29190v1","created_at":"2026-05-29T01:05:23.464890+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.29190","created_at":"2026-05-29T01:05:23.464890+00:00"},{"alias_kind":"pith_short_12","alias_value":"LQB5USZGAMNL","created_at":"2026-05-29T01:05:23.464890+00:00"},{"alias_kind":"pith_short_16","alias_value":"LQB5USZGAMNLLIZ6","created_at":"2026-05-29T01:05:23.464890+00:00"},{"alias_kind":"pith_short_8","alias_value":"LQB5USZG","created_at":"2026-05-29T01:05:23.464890+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LQB5USZGAMNLLIZ6TUUUCV55NL","json":"https://pith.science/pith/LQB5USZGAMNLLIZ6TUUUCV55NL.json","graph_json":"https://pith.science/api/pith-number/LQB5USZGAMNLLIZ6TUUUCV55NL/graph.json","events_json":"https://pith.science/api/pith-number/LQB5USZGAMNLLIZ6TUUUCV55NL/events.json","paper":"https://pith.science/paper/LQB5USZG"},"agent_actions":{"view_html":"https://pith.science/pith/LQB5USZGAMNLLIZ6TUUUCV55NL","download_json":"https://pith.science/pith/LQB5USZGAMNLLIZ6TUUUCV55NL.json","view_paper":"https://pith.science/paper/LQB5USZG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.29190&json=true","fetch_graph":"https://pith.science/api/pith-number/LQB5USZGAMNLLIZ6TUUUCV55NL/graph.json","fetch_events":"https://pith.science/api/pith-number/LQB5USZGAMNLLIZ6TUUUCV55NL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LQB5USZGAMNLLIZ6TUUUCV55NL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LQB5USZGAMNLLIZ6TUUUCV55NL/action/storage_attestation","attest_author":"https://pith.science/pith/LQB5USZGAMNLLIZ6TUUUCV55NL/action/author_attestation","sign_citation":"https://pith.science/pith/LQB5USZGAMNLLIZ6TUUUCV55NL/action/citation_signature","submit_replication":"https://pith.science/pith/LQB5USZGAMNLLIZ6TUUUCV55NL/action/replication_record"}},"created_at":"2026-05-29T01:05:23.464890+00:00","updated_at":"2026-05-29T01:05:23.464890+00:00"}