{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:GQHJ7K7WHXUWQKNNMCJCJFQHVL","short_pith_number":"pith:GQHJ7K7W","canonical_record":{"source":{"id":"2603.16140","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-03-17T05:48:32Z","cross_cats_sorted":[],"title_canon_sha256":"31371ab442676772e04079c9d630f1250c4453db79d2ee0e8176a11661227007","abstract_canon_sha256":"ab28c2a724e32205381e79c70bd4c178c3f8e638e9e888cd89f17cc0ae63fd53"},"schema_version":"1.0"},"canonical_sha256":"340e9fabf63de96829ad6092249607aae2665a984de75415eedab67c4ea74434","source":{"kind":"arxiv","id":"2603.16140","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.16140","created_at":"2026-07-31T01:10:32Z"},{"alias_kind":"arxiv_version","alias_value":"2603.16140v2","created_at":"2026-07-31T01:10:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.16140","created_at":"2026-07-31T01:10:32Z"},{"alias_kind":"pith_short_12","alias_value":"GQHJ7K7WHXUW","created_at":"2026-07-31T01:10:32Z"},{"alias_kind":"pith_short_16","alias_value":"GQHJ7K7WHXUWQKNN","created_at":"2026-07-31T01:10:32Z"},{"alias_kind":"pith_short_8","alias_value":"GQHJ7K7W","created_at":"2026-07-31T01:10:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:GQHJ7K7WHXUWQKNNMCJCJFQHVL","target":"record","payload":{"canonical_record":{"source":{"id":"2603.16140","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-03-17T05:48:32Z","cross_cats_sorted":[],"title_canon_sha256":"31371ab442676772e04079c9d630f1250c4453db79d2ee0e8176a11661227007","abstract_canon_sha256":"ab28c2a724e32205381e79c70bd4c178c3f8e638e9e888cd89f17cc0ae63fd53"},"schema_version":"1.0"},"canonical_sha256":"340e9fabf63de96829ad6092249607aae2665a984de75415eedab67c4ea74434","receipt":{"kind":"pith_receipt","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"340e9fabf63de96829ad6092249607aae2665a984de75415eedab67c4ea74434","last_reissued_at":"2026-07-31T01:10:32.275724Z","signature_status":"unsigned_v0","first_computed_at":"2026-07-31T01:10:32.275724Z"},"source_kind":"arxiv","source_id":"2603.16140","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-31T01:10:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XXdazycMX2DJQ7lltGaGhf/67ycKHE9WGKztakN/zwdBrCWH9hW5xeU/oIoeuLoJNJ97cxqZe+S3zZPy34syAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T00:28:12.911219Z"},"content_sha256":"bef2e7b29666e71f630624ca5eb9950be6a9617a1955570105deade94c908329","schema_version":"1.0","event_id":"sha256:bef2e7b29666e71f630624ca5eb9950be6a9617a1955570105deade94c908329"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:GQHJ7K7WHXUWQKNNMCJCJFQHVL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Noisy Data is Destructive to Reinforcement Learning with Verifiable Rewards","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Daniel Kang, Yuxuan Zhu","submitted_at":"2026-03-17T05:48:32Z","abstract_excerpt":"Reinforcement learning with verifiable rewards (RLVR) has driven recent capability advances of large language models across various domains. Recent studies suggest that improved RLVR algorithms allow models to learn effectively from incorrect annotations, achieving performance comparable to learning from clean data. In this work, we show that these findings are invalid because the claimed 100% noisy training data is \"contaminated\" with clean data. After rectifying the dataset with a rigorous re-verification pipeline, we demonstrate that noise is destructive to RLVR. We show that existing RLVR "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.16140","kind":"arxiv","version":2},"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/2603.16140/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":"2026-07-31T01:10:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"30CO6Dkwi/X3fLohh4Esdv7giRIQ9xsR6osWH56UP0YHJPyaYoo7CYBkBkaVfMWo9VIoJbyyNdTpsrsXLJjjCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T00:28:12.911712Z"},"content_sha256":"9b1229894f4edcc94c04b8a66c53e348b2a935748c8155a4fc4523bd5e24facc","schema_version":"1.0","event_id":"sha256:9b1229894f4edcc94c04b8a66c53e348b2a935748c8155a4fc4523bd5e24facc"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/GQHJ7K7WHXUWQKNNMCJCJFQHVL/bundle.json","state_url":"https://pith.science/pith/GQHJ7K7WHXUWQKNNMCJCJFQHVL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/GQHJ7K7WHXUWQKNNMCJCJFQHVL/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":"2026-08-04T00:28:12Z","links":{"resolver":"https://pith.science/pith/GQHJ7K7WHXUWQKNNMCJCJFQHVL","bundle":"https://pith.science/pith/GQHJ7K7WHXUWQKNNMCJCJFQHVL/bundle.json","state":"https://pith.science/pith/GQHJ7K7WHXUWQKNNMCJCJFQHVL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/GQHJ7K7WHXUWQKNNMCJCJFQHVL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:GQHJ7K7WHXUWQKNNMCJCJFQHVL","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":"ab28c2a724e32205381e79c70bd4c178c3f8e638e9e888cd89f17cc0ae63fd53","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-03-17T05:48:32Z","title_canon_sha256":"31371ab442676772e04079c9d630f1250c4453db79d2ee0e8176a11661227007"},"schema_version":"1.0","source":{"id":"2603.16140","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.16140","created_at":"2026-07-31T01:10:32Z"},{"alias_kind":"arxiv_version","alias_value":"2603.16140v2","created_at":"2026-07-31T01:10:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.16140","created_at":"2026-07-31T01:10:32Z"},{"alias_kind":"pith_short_12","alias_value":"GQHJ7K7WHXUW","created_at":"2026-07-31T01:10:32Z"},{"alias_kind":"pith_short_16","alias_value":"GQHJ7K7WHXUWQKNN","created_at":"2026-07-31T01:10:32Z"},{"alias_kind":"pith_short_8","alias_value":"GQHJ7K7W","created_at":"2026-07-31T01:10:32Z"}],"graph_snapshots":[{"event_id":"sha256:9b1229894f4edcc94c04b8a66c53e348b2a935748c8155a4fc4523bd5e24facc","target":"graph","created_at":"2026-07-31T01:10:32Z","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/2603.16140/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement learning with verifiable rewards (RLVR) has driven recent capability advances of large language models across various domains. Recent studies suggest that improved RLVR algorithms allow models to learn effectively from incorrect annotations, achieving performance comparable to learning from clean data. In this work, we show that these findings are invalid because the claimed 100% noisy training data is \"contaminated\" with clean data. After rectifying the dataset with a rigorous re-verification pipeline, we demonstrate that noise is destructive to RLVR. We show that existing RLVR ","authors_text":"Daniel Kang, Yuxuan Zhu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-03-17T05:48:32Z","title":"Noisy Data is Destructive to Reinforcement Learning with Verifiable Rewards"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.16140","kind":"arxiv","version":2},"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:bef2e7b29666e71f630624ca5eb9950be6a9617a1955570105deade94c908329","target":"record","created_at":"2026-07-31T01:10:32Z","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":"ab28c2a724e32205381e79c70bd4c178c3f8e638e9e888cd89f17cc0ae63fd53","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-03-17T05:48:32Z","title_canon_sha256":"31371ab442676772e04079c9d630f1250c4453db79d2ee0e8176a11661227007"},"schema_version":"1.0","source":{"id":"2603.16140","kind":"arxiv","version":2}},"canonical_sha256":"340e9fabf63de96829ad6092249607aae2665a984de75415eedab67c4ea74434","receipt":{"builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"340e9fabf63de96829ad6092249607aae2665a984de75415eedab67c4ea74434","first_computed_at":"2026-07-31T01:10:32.275724Z","kind":"pith_receipt","last_reissued_at":"2026-07-31T01:10:32.275724Z","receipt_version":"0.3","signature_status":"unsigned_v0"},"source_id":"2603.16140","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bef2e7b29666e71f630624ca5eb9950be6a9617a1955570105deade94c908329","sha256:9b1229894f4edcc94c04b8a66c53e348b2a935748c8155a4fc4523bd5e24facc"],"state_sha256":"50b5fa77b6cb1e9265fcb88e48d69e2db638308a068f6e258e0cd80b72cf4c54"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"EcIttoP2A/HqVDopIZ1P6UG9UzdmaX1dDChYNtT0D3Yk9mFPC0/QIwEfjFoQB+UN9C7fuwp6Xr560bHSa3ZgCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T00:28:12.916276Z","bundle_sha256":"4f9d9d491ecc4b056184e818ad9b45f25471894986b6fb17e32323fb68aa6057"}}