{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:X2AL22FNCIXDXXZWZHZLTAYGJ4","short_pith_number":"pith:X2AL22FN","canonical_record":{"source":{"id":"2407.18248","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-25T17:59:16Z","cross_cats_sorted":[],"title_canon_sha256":"d211a5a05cadc64e967203bba451412e37d3c9a13b44f72a8b22a365a5eb2ae2","abstract_canon_sha256":"1704984e80264b625dde1978d68ff01309cf2fafdd00d7bc7bdd8de829bd1cea"},"schema_version":"1.0"},"canonical_sha256":"be80bd68ad122e3bdf36c9f2b983064f2efb95ae643f372167ac653842210189","source":{"kind":"arxiv","id":"2407.18248","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.18248","created_at":"2026-07-05T08:48:36Z"},{"alias_kind":"arxiv_version","alias_value":"2407.18248v1","created_at":"2026-07-05T08:48:36Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.18248","created_at":"2026-07-05T08:48:36Z"},{"alias_kind":"pith_short_12","alias_value":"X2AL22FNCIXD","created_at":"2026-07-05T08:48:36Z"},{"alias_kind":"pith_short_16","alias_value":"X2AL22FNCIXDXXZW","created_at":"2026-07-05T08:48:36Z"},{"alias_kind":"pith_short_8","alias_value":"X2AL22FN","created_at":"2026-07-05T08:48:36Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:X2AL22FNCIXDXXZWZHZLTAYGJ4","target":"record","payload":{"canonical_record":{"source":{"id":"2407.18248","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-25T17:59:16Z","cross_cats_sorted":[],"title_canon_sha256":"d211a5a05cadc64e967203bba451412e37d3c9a13b44f72a8b22a365a5eb2ae2","abstract_canon_sha256":"1704984e80264b625dde1978d68ff01309cf2fafdd00d7bc7bdd8de829bd1cea"},"schema_version":"1.0"},"canonical_sha256":"be80bd68ad122e3bdf36c9f2b983064f2efb95ae643f372167ac653842210189","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:48:36.624225Z","signature_b64":"i4Um2vtppJPQ9kWVsv+PpCTmf9RW0lk0YL4NOIBYw4EuX1pGBcoSssQX2w/v8DgMno8xLd48jClTBNVt2sSUDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"be80bd68ad122e3bdf36c9f2b983064f2efb95ae643f372167ac653842210189","last_reissued_at":"2026-07-05T08:48:36.623732Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:48:36.623732Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2407.18248","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":"2026-07-05T08:48:36Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9AftDHoj057UPCdXxNm5WdQj+Stsq5c3xNmvrjYPY6IIW2K0CrH369XEyX7SYDPR67acCO+f6/Sexb8qZ8oeBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T23:37:58.997810Z"},"content_sha256":"7691d19c64ca7e40fa5e6532ae1888a40d57be28e67ab12fcdabbafbe2ce7004","schema_version":"1.0","event_id":"sha256:7691d19c64ca7e40fa5e6532ae1888a40d57be28e67ab12fcdabbafbe2ce7004"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:X2AL22FNCIXDXXZWZHZLTAYGJ4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Self-Training with Direct Preference Optimization Improves Chain-of-Thought Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Shichen Li, Tianduo Wang, Wei Lu","submitted_at":"2024-07-25T17:59:16Z","abstract_excerpt":"Effective training of language models (LMs) for mathematical reasoning tasks demands high-quality supervised fine-tuning data. Besides obtaining annotations from human experts, a common alternative is sampling from larger and more powerful LMs. However, this knowledge distillation approach can be costly and unstable, particularly when relying on closed-source, proprietary LMs like GPT-4, whose behaviors are often unpredictable. In this work, we demonstrate that the reasoning abilities of small-scale LMs can be enhanced through self-training, a process where models learn from their own outputs."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.18248","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/2407.18248/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":"2026-07-05T08:48:36Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"63VlUV7duj1QxMtSLzfyiolrdoHAzJSzz1y5bkssIYQI4mLzTXVN4tnP0ajKUGJiplKbxLDm5ezBIAv48iRnBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T23:37:58.998327Z"},"content_sha256":"dc8f6a978f911cd29105b343816c76c0aea0ab2f5d12e348d49419679becfcf5","schema_version":"1.0","event_id":"sha256:dc8f6a978f911cd29105b343816c76c0aea0ab2f5d12e348d49419679becfcf5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/X2AL22FNCIXDXXZWZHZLTAYGJ4/bundle.json","state_url":"https://pith.science/pith/X2AL22FNCIXDXXZWZHZLTAYGJ4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/X2AL22FNCIXDXXZWZHZLTAYGJ4/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":"2026-08-08T23:37:59Z","links":{"resolver":"https://pith.science/pith/X2AL22FNCIXDXXZWZHZLTAYGJ4","bundle":"https://pith.science/pith/X2AL22FNCIXDXXZWZHZLTAYGJ4/bundle.json","state":"https://pith.science/pith/X2AL22FNCIXDXXZWZHZLTAYGJ4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/X2AL22FNCIXDXXZWZHZLTAYGJ4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:X2AL22FNCIXDXXZWZHZLTAYGJ4","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":"1704984e80264b625dde1978d68ff01309cf2fafdd00d7bc7bdd8de829bd1cea","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-25T17:59:16Z","title_canon_sha256":"d211a5a05cadc64e967203bba451412e37d3c9a13b44f72a8b22a365a5eb2ae2"},"schema_version":"1.0","source":{"id":"2407.18248","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.18248","created_at":"2026-07-05T08:48:36Z"},{"alias_kind":"arxiv_version","alias_value":"2407.18248v1","created_at":"2026-07-05T08:48:36Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.18248","created_at":"2026-07-05T08:48:36Z"},{"alias_kind":"pith_short_12","alias_value":"X2AL22FNCIXD","created_at":"2026-07-05T08:48:36Z"},{"alias_kind":"pith_short_16","alias_value":"X2AL22FNCIXDXXZW","created_at":"2026-07-05T08:48:36Z"},{"alias_kind":"pith_short_8","alias_value":"X2AL22FN","created_at":"2026-07-05T08:48:36Z"}],"graph_snapshots":[{"event_id":"sha256:dc8f6a978f911cd29105b343816c76c0aea0ab2f5d12e348d49419679becfcf5","target":"graph","created_at":"2026-07-05T08:48:36Z","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/2407.18248/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Effective training of language models (LMs) for mathematical reasoning tasks demands high-quality supervised fine-tuning data. Besides obtaining annotations from human experts, a common alternative is sampling from larger and more powerful LMs. However, this knowledge distillation approach can be costly and unstable, particularly when relying on closed-source, proprietary LMs like GPT-4, whose behaviors are often unpredictable. In this work, we demonstrate that the reasoning abilities of small-scale LMs can be enhanced through self-training, a process where models learn from their own outputs.","authors_text":"Shichen Li, Tianduo Wang, Wei Lu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-25T17:59:16Z","title":"Self-Training with Direct Preference Optimization Improves Chain-of-Thought Reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.18248","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":"sha256:7691d19c64ca7e40fa5e6532ae1888a40d57be28e67ab12fcdabbafbe2ce7004","target":"record","created_at":"2026-07-05T08:48:36Z","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":"1704984e80264b625dde1978d68ff01309cf2fafdd00d7bc7bdd8de829bd1cea","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-25T17:59:16Z","title_canon_sha256":"d211a5a05cadc64e967203bba451412e37d3c9a13b44f72a8b22a365a5eb2ae2"},"schema_version":"1.0","source":{"id":"2407.18248","kind":"arxiv","version":1}},"canonical_sha256":"be80bd68ad122e3bdf36c9f2b983064f2efb95ae643f372167ac653842210189","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"be80bd68ad122e3bdf36c9f2b983064f2efb95ae643f372167ac653842210189","first_computed_at":"2026-07-05T08:48:36.623732Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:48:36.623732Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"i4Um2vtppJPQ9kWVsv+PpCTmf9RW0lk0YL4NOIBYw4EuX1pGBcoSssQX2w/v8DgMno8xLd48jClTBNVt2sSUDw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:48:36.624225Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.18248","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7691d19c64ca7e40fa5e6532ae1888a40d57be28e67ab12fcdabbafbe2ce7004","sha256:dc8f6a978f911cd29105b343816c76c0aea0ab2f5d12e348d49419679becfcf5"],"state_sha256":"380417a615a08fef681ba73531db607dbdbcfd1d760ff4cc4073b126cd9783f1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"unDRia/g4JqLFfTpPr4lp1D3roGNW8/Z4tPiJGkWe2Tzw5G2TMSGd5ExlcfbMX04U+J7Kne2jGSKTadz2QUzDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T23:37:59.004086Z","bundle_sha256":"ca0c39244aa459db33a2e57cafa60a83176f46cac2ddad1f0dab9221300af8a3"}}