{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:EHL5ZXW5YGNEBAZ23JEZN7LLNC","short_pith_number":"pith:EHL5ZXW5","canonical_record":{"source":{"id":"2504.04736","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-04-07T05:20:58Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"56da1d0d60a1c49baca11beaa0c751678a80bc34e9633644153fe5bfa5299b33","abstract_canon_sha256":"120ef739dce2eddded0d53efed0828c56cc144277d9d049f079eeb893c1327a1"},"schema_version":"1.0"},"canonical_sha256":"21d7dcdeddc19a40833ada4996fd6b68a40dd834216b503dd8034023327826b3","source":{"kind":"arxiv","id":"2504.04736","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.04736","created_at":"2026-07-05T10:54:45Z"},{"alias_kind":"arxiv_version","alias_value":"2504.04736v2","created_at":"2026-07-05T10:54:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.04736","created_at":"2026-07-05T10:54:45Z"},{"alias_kind":"pith_short_12","alias_value":"EHL5ZXW5YGNE","created_at":"2026-07-05T10:54:45Z"},{"alias_kind":"pith_short_16","alias_value":"EHL5ZXW5YGNEBAZ2","created_at":"2026-07-05T10:54:45Z"},{"alias_kind":"pith_short_8","alias_value":"EHL5ZXW5","created_at":"2026-07-05T10:54:45Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:EHL5ZXW5YGNEBAZ23JEZN7LLNC","target":"record","payload":{"canonical_record":{"source":{"id":"2504.04736","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-04-07T05:20:58Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"56da1d0d60a1c49baca11beaa0c751678a80bc34e9633644153fe5bfa5299b33","abstract_canon_sha256":"120ef739dce2eddded0d53efed0828c56cc144277d9d049f079eeb893c1327a1"},"schema_version":"1.0"},"canonical_sha256":"21d7dcdeddc19a40833ada4996fd6b68a40dd834216b503dd8034023327826b3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:54:45.891178Z","signature_b64":"lXu0K+9KLFh9+n+tzjgAFRDLcbIvvRKtfRxn2DlHvglpftHe6GXIJ7rUDsSYLN+UHv/IIxegQ4EH8t7J2Yw2Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"21d7dcdeddc19a40833ada4996fd6b68a40dd834216b503dd8034023327826b3","last_reissued_at":"2026-07-05T10:54:45.890672Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:54:45.890672Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2504.04736","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":"2026-07-05T10:54:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Mwmy2F8l56IK7nofbypM4njqH8PBTLKyQcIEYTPWrcXPSgcAsZ5dszBZz0FcHRjwsqT2mU3CNHX29Zq+uA3PDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T00:03:11.499305Z"},"content_sha256":"16db682d4bab53f05e79e4694f15ce364e413ca763b5234c3bfe4cb759bc6be3","schema_version":"1.0","event_id":"sha256:16db682d4bab53f05e79e4694f15ce364e413ca763b5234c3bfe4cb759bc6be3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:EHL5ZXW5YGNEBAZ23JEZN7LLNC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Synthetic Data Generation & Multi-Step RL for Reasoning & Tool Use","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Anna Goldie, Azalia Mirhoseini, Christopher D. Manning, Hao Zhou, Irene Cai","submitted_at":"2025-04-07T05:20:58Z","abstract_excerpt":"Reinforcement learning has been shown to improve the performance of large language models. However, traditional approaches like RLHF or RLAIF treat the problem as single-step. As focus shifts toward more complex reasoning and agentic tasks, language models must take multiple steps of text generation, reasoning and environment interaction before generating a solution. We propose a synthetic data generation and RL methodology targeting multi-step optimization scenarios. This approach, called Step-Wise Reinforcement Learning (SWiRL), iteratively generates multi-step reasoning and tool use data, a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.04736","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/2504.04736/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":"2026-07-05T10:54:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rWVRJqD6OerM04Dns/1Wek1TFmMjUgXilWqnUHVq7rdTz967WVAPAOKVrgnREvCg05WgdxY/tPfCCpIUIsJgCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T00:03:11.499817Z"},"content_sha256":"63c1424a43a292f5fc297396846262de7a89a70786deafa99245d5240d92032c","schema_version":"1.0","event_id":"sha256:63c1424a43a292f5fc297396846262de7a89a70786deafa99245d5240d92032c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/EHL5ZXW5YGNEBAZ23JEZN7LLNC/bundle.json","state_url":"https://pith.science/pith/EHL5ZXW5YGNEBAZ23JEZN7LLNC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/EHL5ZXW5YGNEBAZ23JEZN7LLNC/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":"2026-08-04T00:03:11Z","links":{"resolver":"https://pith.science/pith/EHL5ZXW5YGNEBAZ23JEZN7LLNC","bundle":"https://pith.science/pith/EHL5ZXW5YGNEBAZ23JEZN7LLNC/bundle.json","state":"https://pith.science/pith/EHL5ZXW5YGNEBAZ23JEZN7LLNC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/EHL5ZXW5YGNEBAZ23JEZN7LLNC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:EHL5ZXW5YGNEBAZ23JEZN7LLNC","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":"120ef739dce2eddded0d53efed0828c56cc144277d9d049f079eeb893c1327a1","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-04-07T05:20:58Z","title_canon_sha256":"56da1d0d60a1c49baca11beaa0c751678a80bc34e9633644153fe5bfa5299b33"},"schema_version":"1.0","source":{"id":"2504.04736","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2504.04736","created_at":"2026-07-05T10:54:45Z"},{"alias_kind":"arxiv_version","alias_value":"2504.04736v2","created_at":"2026-07-05T10:54:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.04736","created_at":"2026-07-05T10:54:45Z"},{"alias_kind":"pith_short_12","alias_value":"EHL5ZXW5YGNE","created_at":"2026-07-05T10:54:45Z"},{"alias_kind":"pith_short_16","alias_value":"EHL5ZXW5YGNEBAZ2","created_at":"2026-07-05T10:54:45Z"},{"alias_kind":"pith_short_8","alias_value":"EHL5ZXW5","created_at":"2026-07-05T10:54:45Z"}],"graph_snapshots":[{"event_id":"sha256:63c1424a43a292f5fc297396846262de7a89a70786deafa99245d5240d92032c","target":"graph","created_at":"2026-07-05T10:54:45Z","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/2504.04736/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement learning has been shown to improve the performance of large language models. However, traditional approaches like RLHF or RLAIF treat the problem as single-step. As focus shifts toward more complex reasoning and agentic tasks, language models must take multiple steps of text generation, reasoning and environment interaction before generating a solution. We propose a synthetic data generation and RL methodology targeting multi-step optimization scenarios. This approach, called Step-Wise Reinforcement Learning (SWiRL), iteratively generates multi-step reasoning and tool use data, a","authors_text":"Anna Goldie, Azalia Mirhoseini, Christopher D. Manning, Hao Zhou, Irene Cai","cross_cats":["cs.CL","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-04-07T05:20:58Z","title":"Synthetic Data Generation & Multi-Step RL for Reasoning & Tool Use"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.04736","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":"sha256:16db682d4bab53f05e79e4694f15ce364e413ca763b5234c3bfe4cb759bc6be3","target":"record","created_at":"2026-07-05T10:54:45Z","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":"120ef739dce2eddded0d53efed0828c56cc144277d9d049f079eeb893c1327a1","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-04-07T05:20:58Z","title_canon_sha256":"56da1d0d60a1c49baca11beaa0c751678a80bc34e9633644153fe5bfa5299b33"},"schema_version":"1.0","source":{"id":"2504.04736","kind":"arxiv","version":2}},"canonical_sha256":"21d7dcdeddc19a40833ada4996fd6b68a40dd834216b503dd8034023327826b3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"21d7dcdeddc19a40833ada4996fd6b68a40dd834216b503dd8034023327826b3","first_computed_at":"2026-07-05T10:54:45.890672Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:54:45.890672Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"lXu0K+9KLFh9+n+tzjgAFRDLcbIvvRKtfRxn2DlHvglpftHe6GXIJ7rUDsSYLN+UHv/IIxegQ4EH8t7J2Yw2Bw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:54:45.891178Z","signed_message":"canonical_sha256_bytes"},"source_id":"2504.04736","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:16db682d4bab53f05e79e4694f15ce364e413ca763b5234c3bfe4cb759bc6be3","sha256:63c1424a43a292f5fc297396846262de7a89a70786deafa99245d5240d92032c"],"state_sha256":"e56addc4f7fe2ad8d7ceca769d51e847f0135b1580583bb85a9b138157739968"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Q2jTh1JpOTc3O7nnZctdgXJ2Tbh3kYz/rbqIr8xdyXzs0XOBUuPjhJpRyI1dZ/RNqRcT+C0ElqfctnMHrtItBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T00:03:11.503572Z","bundle_sha256":"caa6069a9f44b4184a43655e1d14eb787bfc96e8bb67a2357c49dccc70baf53b"}}