{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4VV4O2XCSASKLCDWQICMNPZEL3","short_pith_number":"pith:4VV4O2XC","schema_version":"1.0","canonical_sha256":"e56bc76ae29024a588768204c6bf245eee47ef87f98ce475fcb21c32bee10fdd","source":{"kind":"arxiv","id":"2606.10768","version":1},"attestation_state":"computed","paper":{"title":"N-GRPO: Embedding-Level Neighbor Mixing for Enhanced Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Hang Yu, Linchao Zhu, Peng Di, Xukun Zhu","submitted_at":"2026-06-09T12:21:27Z","abstract_excerpt":"The success of Large Language Models in mathematical reasoning relies heavily on the generation of diverse and valid solution paths during the rollout phase. However, current rollout techniques face a fundamental trade-off: token-level sampling often yields redundant trajectories that differ only in rephrasing, while embedding-level methods utilizing random noise frequently disrupt semantic consistency. To resolve this, we introduce N-GRPO, a novel exploration strategy integrated into the Group Relative Policy Optimization (GRPO) framework. Rather than relying on token-level sampling or native"},"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":"2606.10768","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-09T12:21:27Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"6f348922b1e442f7f4f0ce163e541259b9f9f6e026a1421eacb1c7b5c1f0530d","abstract_canon_sha256":"3fb22966397eed8205841b285770cb14e8022a4bd6e5342e925c56ae7e24da92"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-10T01:10:39.550804Z","signature_b64":"uVBz6yGbM/xzoeGx5mn/ySJbL/tRgFwaml4ZkbgqIAlmZLwcZECGaFYerI7nLIereigPvzkPLLeFx1JmWhFLDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e56bc76ae29024a588768204c6bf245eee47ef87f98ce475fcb21c32bee10fdd","last_reissued_at":"2026-06-10T01:10:39.549840Z","signature_status":"signed_v1","first_computed_at":"2026-06-10T01:10:39.549840Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"N-GRPO: Embedding-Level Neighbor Mixing for Enhanced Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Hang Yu, Linchao Zhu, Peng Di, Xukun Zhu","submitted_at":"2026-06-09T12:21:27Z","abstract_excerpt":"The success of Large Language Models in mathematical reasoning relies heavily on the generation of diverse and valid solution paths during the rollout phase. However, current rollout techniques face a fundamental trade-off: token-level sampling often yields redundant trajectories that differ only in rephrasing, while embedding-level methods utilizing random noise frequently disrupt semantic consistency. To resolve this, we introduce N-GRPO, a novel exploration strategy integrated into the Group Relative Policy Optimization (GRPO) framework. Rather than relying on token-level sampling or native"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.10768","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/2606.10768/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":"2606.10768","created_at":"2026-06-10T01:10:39.550011+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.10768v1","created_at":"2026-06-10T01:10:39.550011+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.10768","created_at":"2026-06-10T01:10:39.550011+00:00"},{"alias_kind":"pith_short_12","alias_value":"4VV4O2XCSASK","created_at":"2026-06-10T01:10:39.550011+00:00"},{"alias_kind":"pith_short_16","alias_value":"4VV4O2XCSASKLCDW","created_at":"2026-06-10T01:10:39.550011+00:00"},{"alias_kind":"pith_short_8","alias_value":"4VV4O2XC","created_at":"2026-06-10T01:10:39.550011+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/4VV4O2XCSASKLCDWQICMNPZEL3","json":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3.json","graph_json":"https://pith.science/api/pith-number/4VV4O2XCSASKLCDWQICMNPZEL3/graph.json","events_json":"https://pith.science/api/pith-number/4VV4O2XCSASKLCDWQICMNPZEL3/events.json","paper":"https://pith.science/paper/4VV4O2XC"},"agent_actions":{"view_html":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3","download_json":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3.json","view_paper":"https://pith.science/paper/4VV4O2XC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.10768&json=true","fetch_graph":"https://pith.science/api/pith-number/4VV4O2XCSASKLCDWQICMNPZEL3/graph.json","fetch_events":"https://pith.science/api/pith-number/4VV4O2XCSASKLCDWQICMNPZEL3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3/action/storage_attestation","attest_author":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3/action/author_attestation","sign_citation":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3/action/citation_signature","submit_replication":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3/action/replication_record"}},"created_at":"2026-06-10T01:10:39.550011+00:00","updated_at":"2026-06-10T01:10:39.550011+00:00"}