{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:4VV4O2XCSASKLCDWQICMNPZEL3","short_pith_number":"pith:4VV4O2XC","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"},"canonical_sha256":"e56bc76ae29024a588768204c6bf245eee47ef87f98ce475fcb21c32bee10fdd","source":{"kind":"arxiv","id":"2606.10768","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.10768","created_at":"2026-06-10T01:10:39Z"},{"alias_kind":"arxiv_version","alias_value":"2606.10768v1","created_at":"2026-06-10T01:10:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.10768","created_at":"2026-06-10T01:10:39Z"},{"alias_kind":"pith_short_12","alias_value":"4VV4O2XCSASK","created_at":"2026-06-10T01:10:39Z"},{"alias_kind":"pith_short_16","alias_value":"4VV4O2XCSASKLCDW","created_at":"2026-06-10T01:10:39Z"},{"alias_kind":"pith_short_8","alias_value":"4VV4O2XC","created_at":"2026-06-10T01:10:39Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:4VV4O2XCSASKLCDWQICMNPZEL3","target":"record","payload":{"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"},"canonical_sha256":"e56bc76ae29024a588768204c6bf245eee47ef87f98ce475fcb21c32bee10fdd","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"},"source_kind":"arxiv","source_id":"2606.10768","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-10T01:10:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kXEJxePexnC88jH1NAkbkSQq/UrG0M4S2d65f1WyJGB3ldENe/nIUwffxAfAM9+uOOKoiy9/8BDPWK80ZuKGDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-10T23:26:15.553637Z"},"content_sha256":"bf48fd7fb97e2fd11549720cb95380c0a42595b1054912b5217652baf5d059c1","schema_version":"1.0","event_id":"sha256:bf48fd7fb97e2fd11549720cb95380c0a42595b1054912b5217652baf5d059c1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:4VV4O2XCSASKLCDWQICMNPZEL3","target":"graph","payload":{"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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-10T01:10:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VFnnUWLkyoCq6aFlNbZmWjAZTHmeNnnBJVuIy1H+rJZwJDYOA8PamdLLDPOf523TnKp4onjk2PCydC9+3J8zBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-10T23:26:15.554457Z"},"content_sha256":"4a9532092c8c244fb9ba387b76e1c547fe2ed036afad42fb79ff0bdab76a1bfb","schema_version":"1.0","event_id":"sha256:4a9532092c8c244fb9ba387b76e1c547fe2ed036afad42fb79ff0bdab76a1bfb"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3/bundle.json","state_url":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/4VV4O2XCSASKLCDWQICMNPZEL3/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-10T23:26:15Z","links":{"resolver":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3","bundle":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3/bundle.json","state":"https://pith.science/pith/4VV4O2XCSASKLCDWQICMNPZEL3/state.json","well_known_bundle":"https://pith.science/.well-known/pith/4VV4O2XCSASKLCDWQICMNPZEL3/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:4VV4O2XCSASKLCDWQICMNPZEL3","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":"3fb22966397eed8205841b285770cb14e8022a4bd6e5342e925c56ae7e24da92","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-09T12:21:27Z","title_canon_sha256":"6f348922b1e442f7f4f0ce163e541259b9f9f6e026a1421eacb1c7b5c1f0530d"},"schema_version":"1.0","source":{"id":"2606.10768","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.10768","created_at":"2026-06-10T01:10:39Z"},{"alias_kind":"arxiv_version","alias_value":"2606.10768v1","created_at":"2026-06-10T01:10:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.10768","created_at":"2026-06-10T01:10:39Z"},{"alias_kind":"pith_short_12","alias_value":"4VV4O2XCSASK","created_at":"2026-06-10T01:10:39Z"},{"alias_kind":"pith_short_16","alias_value":"4VV4O2XCSASKLCDW","created_at":"2026-06-10T01:10:39Z"},{"alias_kind":"pith_short_8","alias_value":"4VV4O2XC","created_at":"2026-06-10T01:10:39Z"}],"graph_snapshots":[{"event_id":"sha256:4a9532092c8c244fb9ba387b76e1c547fe2ed036afad42fb79ff0bdab76a1bfb","target":"graph","created_at":"2026-06-10T01:10:39Z","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/2606.10768/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"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","authors_text":"Hang Yu, Linchao Zhu, Peng Di, Xukun Zhu","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-09T12:21:27Z","title":"N-GRPO: Embedding-Level Neighbor Mixing for Enhanced Policy Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.10768","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:bf48fd7fb97e2fd11549720cb95380c0a42595b1054912b5217652baf5d059c1","target":"record","created_at":"2026-06-10T01:10:39Z","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":"3fb22966397eed8205841b285770cb14e8022a4bd6e5342e925c56ae7e24da92","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-09T12:21:27Z","title_canon_sha256":"6f348922b1e442f7f4f0ce163e541259b9f9f6e026a1421eacb1c7b5c1f0530d"},"schema_version":"1.0","source":{"id":"2606.10768","kind":"arxiv","version":1}},"canonical_sha256":"e56bc76ae29024a588768204c6bf245eee47ef87f98ce475fcb21c32bee10fdd","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e56bc76ae29024a588768204c6bf245eee47ef87f98ce475fcb21c32bee10fdd","first_computed_at":"2026-06-10T01:10:39.549840Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-10T01:10:39.549840Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"uVBz6yGbM/xzoeGx5mn/ySJbL/tRgFwaml4ZkbgqIAlmZLwcZECGaFYerI7nLIereigPvzkPLLeFx1JmWhFLDQ==","signature_status":"signed_v1","signed_at":"2026-06-10T01:10:39.550804Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.10768","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bf48fd7fb97e2fd11549720cb95380c0a42595b1054912b5217652baf5d059c1","sha256:4a9532092c8c244fb9ba387b76e1c547fe2ed036afad42fb79ff0bdab76a1bfb"],"state_sha256":"e87f544e503980b36113422e0da7dfc351565b70baa71c7f475c2aa883c3115b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"1B+ExuX9GUMCPxVyHC4WU2E8R7WZs/zwaRCJgq8XIyEyqWM+vklElnNGgwVThSn7qwF3C8CK7QAOUzLY6VFuCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-10T23:26:15.559232Z","bundle_sha256":"1b86f69b7e877afd0e0f2313c197455d4ccd0883cbd37523dbd87631ffeffa85"}}