{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:A325CT6LKQEKUFICPQKXH62WE3","short_pith_number":"pith:A325CT6L","schema_version":"1.0","canonical_sha256":"06f5d14fcb5408aa15027c1573fb5626cc024c81436c7c7d7418ed1ef92b7cbc","source":{"kind":"arxiv","id":"2606.27369","version":1},"attestation_state":"computed","paper":{"title":"Reinforcement Learning without Ground-Truth Solutions can Improve LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Kun Zhou, Nikki Lijing Kuang, Qiyue Gao, Tongtong Liang, Xunpeng Huang, Yi-An Ma, Yingyu Lin, Yuxiong He, Zhewei Yao","submitted_at":"2026-06-25T17:59:36Z","abstract_excerpt":"Reinforcement learning with verifiable rewards (RLVR) for training LLMs typically rely on ground-truth answers to assign rewards, limiting their applicability to tasks where the ground-truth solution is unknown. We introduce a \\textbf{R}anking-\\textbf{i}nduced \\textbf{VER}ifiable framework (RiVER) that trains LLMs on score-based optimization tasks without ground-truth solutions, using deterministic execution feedback as continuous-valued supervision. When applying group-relative RL to such continuous rewards, we identify two key challenges: \\emph{scale dominance}, where uncalibrated score magn"},"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":"2606.27369","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-06-25T17:59:36Z","cross_cats_sorted":[],"title_canon_sha256":"3ebae9108e9356a8d9026e7d924f18ea579620e823b7b28b146b4231e52a65a9","abstract_canon_sha256":"28db7998274c084c2582d78e4bfd8e69a38833d9ed59ff8cb5e9148af199833f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-26T01:16:20.089413Z","signature_b64":"zdkFbhdIxc8AkBBTIe8EQrPg82UHS0975S5FXClWUnwdjuvodlY0DJsVUxpbDzJ7cCXeHS49hYsABOJY82V6CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"06f5d14fcb5408aa15027c1573fb5626cc024c81436c7c7d7418ed1ef92b7cbc","last_reissued_at":"2026-06-26T01:16:20.089000Z","signature_status":"signed_v1","first_computed_at":"2026-06-26T01:16:20.089000Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reinforcement Learning without Ground-Truth Solutions can Improve LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Kun Zhou, Nikki Lijing Kuang, Qiyue Gao, Tongtong Liang, Xunpeng Huang, Yi-An Ma, Yingyu Lin, Yuxiong He, Zhewei Yao","submitted_at":"2026-06-25T17:59:36Z","abstract_excerpt":"Reinforcement learning with verifiable rewards (RLVR) for training LLMs typically rely on ground-truth answers to assign rewards, limiting their applicability to tasks where the ground-truth solution is unknown. We introduce a \\textbf{R}anking-\\textbf{i}nduced \\textbf{VER}ifiable framework (RiVER) that trains LLMs on score-based optimization tasks without ground-truth solutions, using deterministic execution feedback as continuous-valued supervision. When applying group-relative RL to such continuous rewards, we identify two key challenges: \\emph{scale dominance}, where uncalibrated score magn"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.27369","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/2606.27369/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":"2606.27369","created_at":"2026-06-26T01:16:20.089057+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.27369v1","created_at":"2026-06-26T01:16:20.089057+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.27369","created_at":"2026-06-26T01:16:20.089057+00:00"},{"alias_kind":"pith_short_12","alias_value":"A325CT6LKQEK","created_at":"2026-06-26T01:16:20.089057+00:00"},{"alias_kind":"pith_short_16","alias_value":"A325CT6LKQEKUFIC","created_at":"2026-06-26T01:16:20.089057+00:00"},{"alias_kind":"pith_short_8","alias_value":"A325CT6L","created_at":"2026-06-26T01:16:20.089057+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/A325CT6LKQEKUFICPQKXH62WE3","json":"https://pith.science/pith/A325CT6LKQEKUFICPQKXH62WE3.json","graph_json":"https://pith.science/api/pith-number/A325CT6LKQEKUFICPQKXH62WE3/graph.json","events_json":"https://pith.science/api/pith-number/A325CT6LKQEKUFICPQKXH62WE3/events.json","paper":"https://pith.science/paper/A325CT6L"},"agent_actions":{"view_html":"https://pith.science/pith/A325CT6LKQEKUFICPQKXH62WE3","download_json":"https://pith.science/pith/A325CT6LKQEKUFICPQKXH62WE3.json","view_paper":"https://pith.science/paper/A325CT6L","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.27369&json=true","fetch_graph":"https://pith.science/api/pith-number/A325CT6LKQEKUFICPQKXH62WE3/graph.json","fetch_events":"https://pith.science/api/pith-number/A325CT6LKQEKUFICPQKXH62WE3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/A325CT6LKQEKUFICPQKXH62WE3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/A325CT6LKQEKUFICPQKXH62WE3/action/storage_attestation","attest_author":"https://pith.science/pith/A325CT6LKQEKUFICPQKXH62WE3/action/author_attestation","sign_citation":"https://pith.science/pith/A325CT6LKQEKUFICPQKXH62WE3/action/citation_signature","submit_replication":"https://pith.science/pith/A325CT6LKQEKUFICPQKXH62WE3/action/replication_record"}},"created_at":"2026-06-26T01:16:20.089057+00:00","updated_at":"2026-06-26T01:16:20.089057+00:00"}