{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:YYWX7TX2KTAF5D3C24UEE553XJ","short_pith_number":"pith:YYWX7TX2","canonical_record":{"source":{"id":"2606.22938","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-22T07:16:08Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f6c5bccbaac8597adbb5e9d19431854509765351f415f00810edf6954e40f842","abstract_canon_sha256":"a132852cabe71f5267b4c69e1f0fb7b14a6bae52320d274ef7d44fbc0a1bcaf2"},"schema_version":"1.0"},"canonical_sha256":"c62d7fcefa54c05e8f62d7284277bbba4ccc472eaa7f661c4d7d08eddf55b824","source":{"kind":"arxiv","id":"2606.22938","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.22938","created_at":"2026-06-23T03:14:04Z"},{"alias_kind":"arxiv_version","alias_value":"2606.22938v1","created_at":"2026-06-23T03:14:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.22938","created_at":"2026-06-23T03:14:04Z"},{"alias_kind":"pith_short_12","alias_value":"YYWX7TX2KTAF","created_at":"2026-06-23T03:14:04Z"},{"alias_kind":"pith_short_16","alias_value":"YYWX7TX2KTAF5D3C","created_at":"2026-06-23T03:14:04Z"},{"alias_kind":"pith_short_8","alias_value":"YYWX7TX2","created_at":"2026-06-23T03:14:04Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:YYWX7TX2KTAF5D3C24UEE553XJ","target":"record","payload":{"canonical_record":{"source":{"id":"2606.22938","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-22T07:16:08Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f6c5bccbaac8597adbb5e9d19431854509765351f415f00810edf6954e40f842","abstract_canon_sha256":"a132852cabe71f5267b4c69e1f0fb7b14a6bae52320d274ef7d44fbc0a1bcaf2"},"schema_version":"1.0"},"canonical_sha256":"c62d7fcefa54c05e8f62d7284277bbba4ccc472eaa7f661c4d7d08eddf55b824","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-23T03:14:04.778937Z","signature_b64":"zP+uUx+Kjvxh2LjYiReARhxeEY39YQHYuVVzfyYmvlT+WZ6WEvd6+aFi+9qMz971aXVOofvwqRQCUspe0kLSBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c62d7fcefa54c05e8f62d7284277bbba4ccc472eaa7f661c4d7d08eddf55b824","last_reissued_at":"2026-06-23T03:14:04.778536Z","signature_status":"signed_v1","first_computed_at":"2026-06-23T03:14:04.778536Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.22938","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":"2026-06-23T03:14:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Lr1Wd+PPLeg5ryPGi6XWZC/g9IoqkUc2ftx2siyqMC7lH5n0+wl9fDxkomzkGV3GfJr0gYZYIbrHSvNr+XZ8Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-05T01:16:40.261273Z"},"content_sha256":"28eb0f7413182a83d1a9ea7f11ed3fe5d474e937af55e1b85b46d76bc190a698","schema_version":"1.0","event_id":"sha256:28eb0f7413182a83d1a9ea7f11ed3fe5d474e937af55e1b85b46d76bc190a698"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:YYWX7TX2KTAF5D3C24UEE553XJ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Provable Benefits of RLVR over SFT for Reasoning Models: Learning to Backtrack Efficiently","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Juno Kim, Stanley Wei","submitted_at":"2026-06-22T07:16:08Z","abstract_excerpt":"Recent advances in large language models (LLMs) have demonstrated that reinforcement fine-tuning of pretrained base models can lead to significant gains in reasoning performance at inference time. In this work, we theoretically analyze why reinforcement fine-tuning induces better reasoning ability than purely supervised fine-tuning (SFT) methods. We model chain-of-thought (CoT) reasoning as a pathfinding problem on graphs and compare the popular method of reinforcement learning with verifiable rewards (RLVR) against traditional SFT. We prove that SFT, when trained on golden shortest paths with"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.22938","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/2606.22938/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":"2026-06-23T03:14:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"E8mPV0sRb2yMIWqhiRSEPvdNTa+sNRUQsqZTbB3uuJNFaSj3nEN8kyIhJrBqkRRcol2YaxI32RcjJENoJndFDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-05T01:16:40.261790Z"},"content_sha256":"1d5d75ba6400d83cb3bd5096344ece7d0eb238e6af2ef169cc0e1a699dcd17e4","schema_version":"1.0","event_id":"sha256:1d5d75ba6400d83cb3bd5096344ece7d0eb238e6af2ef169cc0e1a699dcd17e4"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/YYWX7TX2KTAF5D3C24UEE553XJ/bundle.json","state_url":"https://pith.science/pith/YYWX7TX2KTAF5D3C24UEE553XJ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/YYWX7TX2KTAF5D3C24UEE553XJ/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":"2026-07-05T01:16:40Z","links":{"resolver":"https://pith.science/pith/YYWX7TX2KTAF5D3C24UEE553XJ","bundle":"https://pith.science/pith/YYWX7TX2KTAF5D3C24UEE553XJ/bundle.json","state":"https://pith.science/pith/YYWX7TX2KTAF5D3C24UEE553XJ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/YYWX7TX2KTAF5D3C24UEE553XJ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:YYWX7TX2KTAF5D3C24UEE553XJ","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":"a132852cabe71f5267b4c69e1f0fb7b14a6bae52320d274ef7d44fbc0a1bcaf2","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-22T07:16:08Z","title_canon_sha256":"f6c5bccbaac8597adbb5e9d19431854509765351f415f00810edf6954e40f842"},"schema_version":"1.0","source":{"id":"2606.22938","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.22938","created_at":"2026-06-23T03:14:04Z"},{"alias_kind":"arxiv_version","alias_value":"2606.22938v1","created_at":"2026-06-23T03:14:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.22938","created_at":"2026-06-23T03:14:04Z"},{"alias_kind":"pith_short_12","alias_value":"YYWX7TX2KTAF","created_at":"2026-06-23T03:14:04Z"},{"alias_kind":"pith_short_16","alias_value":"YYWX7TX2KTAF5D3C","created_at":"2026-06-23T03:14:04Z"},{"alias_kind":"pith_short_8","alias_value":"YYWX7TX2","created_at":"2026-06-23T03:14:04Z"}],"graph_snapshots":[{"event_id":"sha256:1d5d75ba6400d83cb3bd5096344ece7d0eb238e6af2ef169cc0e1a699dcd17e4","target":"graph","created_at":"2026-06-23T03:14:04Z","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/2606.22938/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent advances in large language models (LLMs) have demonstrated that reinforcement fine-tuning of pretrained base models can lead to significant gains in reasoning performance at inference time. In this work, we theoretically analyze why reinforcement fine-tuning induces better reasoning ability than purely supervised fine-tuning (SFT) methods. We model chain-of-thought (CoT) reasoning as a pathfinding problem on graphs and compare the popular method of reinforcement learning with verifiable rewards (RLVR) against traditional SFT. We prove that SFT, when trained on golden shortest paths with","authors_text":"Juno Kim, Stanley Wei","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-22T07:16:08Z","title":"Provable Benefits of RLVR over SFT for Reasoning Models: Learning to Backtrack Efficiently"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.22938","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":"sha256:28eb0f7413182a83d1a9ea7f11ed3fe5d474e937af55e1b85b46d76bc190a698","target":"record","created_at":"2026-06-23T03:14:04Z","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":"a132852cabe71f5267b4c69e1f0fb7b14a6bae52320d274ef7d44fbc0a1bcaf2","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-06-22T07:16:08Z","title_canon_sha256":"f6c5bccbaac8597adbb5e9d19431854509765351f415f00810edf6954e40f842"},"schema_version":"1.0","source":{"id":"2606.22938","kind":"arxiv","version":1}},"canonical_sha256":"c62d7fcefa54c05e8f62d7284277bbba4ccc472eaa7f661c4d7d08eddf55b824","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c62d7fcefa54c05e8f62d7284277bbba4ccc472eaa7f661c4d7d08eddf55b824","first_computed_at":"2026-06-23T03:14:04.778536Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-23T03:14:04.778536Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"zP+uUx+Kjvxh2LjYiReARhxeEY39YQHYuVVzfyYmvlT+WZ6WEvd6+aFi+9qMz971aXVOofvwqRQCUspe0kLSBA==","signature_status":"signed_v1","signed_at":"2026-06-23T03:14:04.778937Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.22938","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:28eb0f7413182a83d1a9ea7f11ed3fe5d474e937af55e1b85b46d76bc190a698","sha256:1d5d75ba6400d83cb3bd5096344ece7d0eb238e6af2ef169cc0e1a699dcd17e4"],"state_sha256":"a5550e92c50e486241f06cbd27945979c0347b06d025d3a8991652a7247c3f08"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wjWqHTUvAmAgOLya7nRPpddqjZEIUDuXmDkfG9kXM2eJwmZrMCsKnanRGKrdLSqq5Ucr1vTG16Hd7RWybsHYBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-05T01:16:40.263893Z","bundle_sha256":"9c821269a0f47bfd4d804953e1e01f7ca696528445a02b0e548eb884b0510a8f"}}