{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:AB5U4DTNRN7QCYHN5UNZWE4PR4","short_pith_number":"pith:AB5U4DTN","schema_version":"1.0","canonical_sha256":"007b4e0e6d8b7f0160eded1b9b138f8f09f4c84269cf4edd941e7665d7cbccb1","source":{"kind":"arxiv","id":"2605.28631","version":1},"attestation_state":"computed","paper":{"title":"Single-Rollout Hidden-State Dynamics for Training-Free RLVR Data Selection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Daniel F. Schmidt, Jianfei Cai, Jianghao Wu, Jin Ye, Weiqiang Wang, Yasmeen George","submitted_at":"2026-05-27T15:38:09Z","abstract_excerpt":"Reinforcement learning with verifiable rewards (RLVR) can yield large reasoning gains from very few training instances, yet its strong sensitivity to which instances are used makes data selection a central bottleneck. Most existing selection pipelines rely on training-time optimization signals and/or require access to verifiable rewards or ground-truth answers over large candidate pools, which is costly and often infeasible in specialized domains. We study RLVR data selection in a setting where selection must be performed before any RL training and without labels or reward evaluation on the fu"},"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":"2605.28631","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-27T15:38:09Z","cross_cats_sorted":[],"title_canon_sha256":"0dac031a26eea15924f6194a61745982a64346e3eaaea0bd8fa8bcf901ed5296","abstract_canon_sha256":"666cfcdfd3a130edfec5623c21c7e4b29515d3d1bc946c0d6f97b2f0c94c4b5f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-28T02:04:58.351890Z","signature_b64":"0r9ZpI0yVVP1lmEeXeGJXhD9YhZGOOlRS5NJBkQMmmOeOyl32Btk1CS615g/sZ1s9a6Djy4fIxPNYogc9DU5Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"007b4e0e6d8b7f0160eded1b9b138f8f09f4c84269cf4edd941e7665d7cbccb1","last_reissued_at":"2026-05-28T02:04:58.351496Z","signature_status":"signed_v1","first_computed_at":"2026-05-28T02:04:58.351496Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Single-Rollout Hidden-State Dynamics for Training-Free RLVR Data Selection","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Daniel F. Schmidt, Jianfei Cai, Jianghao Wu, Jin Ye, Weiqiang Wang, Yasmeen George","submitted_at":"2026-05-27T15:38:09Z","abstract_excerpt":"Reinforcement learning with verifiable rewards (RLVR) can yield large reasoning gains from very few training instances, yet its strong sensitivity to which instances are used makes data selection a central bottleneck. Most existing selection pipelines rely on training-time optimization signals and/or require access to verifiable rewards or ground-truth answers over large candidate pools, which is costly and often infeasible in specialized domains. We study RLVR data selection in a setting where selection must be performed before any RL training and without labels or reward evaluation on the fu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.28631","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/2605.28631/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":"2605.28631","created_at":"2026-05-28T02:04:58.351553+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.28631v1","created_at":"2026-05-28T02:04:58.351553+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.28631","created_at":"2026-05-28T02:04:58.351553+00:00"},{"alias_kind":"pith_short_12","alias_value":"AB5U4DTNRN7Q","created_at":"2026-05-28T02:04:58.351553+00:00"},{"alias_kind":"pith_short_16","alias_value":"AB5U4DTNRN7QCYHN","created_at":"2026-05-28T02:04:58.351553+00:00"},{"alias_kind":"pith_short_8","alias_value":"AB5U4DTN","created_at":"2026-05-28T02:04:58.351553+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/AB5U4DTNRN7QCYHN5UNZWE4PR4","json":"https://pith.science/pith/AB5U4DTNRN7QCYHN5UNZWE4PR4.json","graph_json":"https://pith.science/api/pith-number/AB5U4DTNRN7QCYHN5UNZWE4PR4/graph.json","events_json":"https://pith.science/api/pith-number/AB5U4DTNRN7QCYHN5UNZWE4PR4/events.json","paper":"https://pith.science/paper/AB5U4DTN"},"agent_actions":{"view_html":"https://pith.science/pith/AB5U4DTNRN7QCYHN5UNZWE4PR4","download_json":"https://pith.science/pith/AB5U4DTNRN7QCYHN5UNZWE4PR4.json","view_paper":"https://pith.science/paper/AB5U4DTN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.28631&json=true","fetch_graph":"https://pith.science/api/pith-number/AB5U4DTNRN7QCYHN5UNZWE4PR4/graph.json","fetch_events":"https://pith.science/api/pith-number/AB5U4DTNRN7QCYHN5UNZWE4PR4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AB5U4DTNRN7QCYHN5UNZWE4PR4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AB5U4DTNRN7QCYHN5UNZWE4PR4/action/storage_attestation","attest_author":"https://pith.science/pith/AB5U4DTNRN7QCYHN5UNZWE4PR4/action/author_attestation","sign_citation":"https://pith.science/pith/AB5U4DTNRN7QCYHN5UNZWE4PR4/action/citation_signature","submit_replication":"https://pith.science/pith/AB5U4DTNRN7QCYHN5UNZWE4PR4/action/replication_record"}},"created_at":"2026-05-28T02:04:58.351553+00:00","updated_at":"2026-05-28T02:04:58.351553+00:00"}