{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:3ISFIZ7QIOMPS3WQU3N7TMG5KC","short_pith_number":"pith:3ISFIZ7Q","canonical_record":{"source":{"id":"2406.10858","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-16T09:06:17Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"40bc6e6e4306b05eeefed9528720c4dcc8242358ae99308a7be80fabc9a0deb4","abstract_canon_sha256":"fdf67f015bcae8d98410a63583a3933e61b5111c58b209871caca7608ff74b67"},"schema_version":"1.0"},"canonical_sha256":"da245467f04398f96ed0a6dbf9b0dd50bad7df6e43cb9cfb9ea529d6ca54e2d2","source":{"kind":"arxiv","id":"2406.10858","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.10858","created_at":"2026-07-05T09:12:29Z"},{"alias_kind":"arxiv_version","alias_value":"2406.10858v2","created_at":"2026-07-05T09:12:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.10858","created_at":"2026-07-05T09:12:29Z"},{"alias_kind":"pith_short_12","alias_value":"3ISFIZ7QIOMP","created_at":"2026-07-05T09:12:29Z"},{"alias_kind":"pith_short_16","alias_value":"3ISFIZ7QIOMPS3WQ","created_at":"2026-07-05T09:12:29Z"},{"alias_kind":"pith_short_8","alias_value":"3ISFIZ7Q","created_at":"2026-07-05T09:12:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:3ISFIZ7QIOMPS3WQU3N7TMG5KC","target":"record","payload":{"canonical_record":{"source":{"id":"2406.10858","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-16T09:06:17Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"40bc6e6e4306b05eeefed9528720c4dcc8242358ae99308a7be80fabc9a0deb4","abstract_canon_sha256":"fdf67f015bcae8d98410a63583a3933e61b5111c58b209871caca7608ff74b67"},"schema_version":"1.0"},"canonical_sha256":"da245467f04398f96ed0a6dbf9b0dd50bad7df6e43cb9cfb9ea529d6ca54e2d2","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:12:29.015482Z","signature_b64":"qukGFD8763m/qVbAhEjroYoo1liq3c/OcOyhXZUyb+0GJ7U+xbk6Gb9cMD00FnAFWxrFjHsbZGwBppaqnb/6AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"da245467f04398f96ed0a6dbf9b0dd50bad7df6e43cb9cfb9ea529d6ca54e2d2","last_reissued_at":"2026-07-05T09:12:29.014980Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:12:29.014980Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.10858","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":"2026-07-05T09:12:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2eb7gDm+M+gGWrN48hWyiD5o2MpNIEUtSInnD6fz3f0ShC8DtW9MWfuANMhDu9woL05kssPFXWeFyr3ePea7DA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T08:38:28.365918Z"},"content_sha256":"4f20004767d41571d9d6d9ae7879c38264b27d4c6520fa3ffc52a6e6ce9cf0e3","schema_version":"1.0","event_id":"sha256:4f20004767d41571d9d6d9ae7879c38264b27d4c6520fa3ffc52a6e6ce9cf0e3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:3ISFIZ7QIOMPS3WQU3N7TMG5KC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Step-level Value Preference Optimization for Mathematical Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chengxi Li, Guoxin Chen, Kai Fan, Minpeng Liao","submitted_at":"2024-06-16T09:06:17Z","abstract_excerpt":"Direct Preference Optimization (DPO) using an implicit reward model has proven to be an effective alternative to reinforcement learning from human feedback (RLHF) for fine-tuning preference aligned large language models (LLMs). However, the overall preference annotations of responses do not fully capture the fine-grained quality of model outputs in complex multi-step reasoning tasks, such as mathematical reasoning. To address this limitation, we introduce a novel algorithm called Step-level Value Preference Optimization (SVPO). Our approach employs Monte Carlo Tree Search (MCTS) to automatical"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.10858","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/2406.10858/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":"2026-07-05T09:12:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HtmJOMdliya0pcqmIfJV86npu7grQe4L1SOWLHk9NJkEZYKmUTJYuDHqK/XaPqnTSCOgl8Y84CBaCaYCpRVHCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T08:38:28.366559Z"},"content_sha256":"fadb94449e18900169cbe7c0c3c8bc56b8e48e4c8ab17c2cace98da748ee9fe1","schema_version":"1.0","event_id":"sha256:fadb94449e18900169cbe7c0c3c8bc56b8e48e4c8ab17c2cace98da748ee9fe1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3ISFIZ7QIOMPS3WQU3N7TMG5KC/bundle.json","state_url":"https://pith.science/pith/3ISFIZ7QIOMPS3WQU3N7TMG5KC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3ISFIZ7QIOMPS3WQU3N7TMG5KC/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":"2026-08-05T08:38:28Z","links":{"resolver":"https://pith.science/pith/3ISFIZ7QIOMPS3WQU3N7TMG5KC","bundle":"https://pith.science/pith/3ISFIZ7QIOMPS3WQU3N7TMG5KC/bundle.json","state":"https://pith.science/pith/3ISFIZ7QIOMPS3WQU3N7TMG5KC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3ISFIZ7QIOMPS3WQU3N7TMG5KC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:3ISFIZ7QIOMPS3WQU3N7TMG5KC","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":"fdf67f015bcae8d98410a63583a3933e61b5111c58b209871caca7608ff74b67","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-16T09:06:17Z","title_canon_sha256":"40bc6e6e4306b05eeefed9528720c4dcc8242358ae99308a7be80fabc9a0deb4"},"schema_version":"1.0","source":{"id":"2406.10858","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.10858","created_at":"2026-07-05T09:12:29Z"},{"alias_kind":"arxiv_version","alias_value":"2406.10858v2","created_at":"2026-07-05T09:12:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.10858","created_at":"2026-07-05T09:12:29Z"},{"alias_kind":"pith_short_12","alias_value":"3ISFIZ7QIOMP","created_at":"2026-07-05T09:12:29Z"},{"alias_kind":"pith_short_16","alias_value":"3ISFIZ7QIOMPS3WQ","created_at":"2026-07-05T09:12:29Z"},{"alias_kind":"pith_short_8","alias_value":"3ISFIZ7Q","created_at":"2026-07-05T09:12:29Z"}],"graph_snapshots":[{"event_id":"sha256:fadb94449e18900169cbe7c0c3c8bc56b8e48e4c8ab17c2cace98da748ee9fe1","target":"graph","created_at":"2026-07-05T09:12:29Z","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/2406.10858/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Direct Preference Optimization (DPO) using an implicit reward model has proven to be an effective alternative to reinforcement learning from human feedback (RLHF) for fine-tuning preference aligned large language models (LLMs). However, the overall preference annotations of responses do not fully capture the fine-grained quality of model outputs in complex multi-step reasoning tasks, such as mathematical reasoning. To address this limitation, we introduce a novel algorithm called Step-level Value Preference Optimization (SVPO). Our approach employs Monte Carlo Tree Search (MCTS) to automatical","authors_text":"Chengxi Li, Guoxin Chen, Kai Fan, Minpeng Liao","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-16T09:06:17Z","title":"Step-level Value Preference Optimization for Mathematical Reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.10858","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":"sha256:4f20004767d41571d9d6d9ae7879c38264b27d4c6520fa3ffc52a6e6ce9cf0e3","target":"record","created_at":"2026-07-05T09:12:29Z","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":"fdf67f015bcae8d98410a63583a3933e61b5111c58b209871caca7608ff74b67","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-16T09:06:17Z","title_canon_sha256":"40bc6e6e4306b05eeefed9528720c4dcc8242358ae99308a7be80fabc9a0deb4"},"schema_version":"1.0","source":{"id":"2406.10858","kind":"arxiv","version":2}},"canonical_sha256":"da245467f04398f96ed0a6dbf9b0dd50bad7df6e43cb9cfb9ea529d6ca54e2d2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"da245467f04398f96ed0a6dbf9b0dd50bad7df6e43cb9cfb9ea529d6ca54e2d2","first_computed_at":"2026-07-05T09:12:29.014980Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:12:29.014980Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"qukGFD8763m/qVbAhEjroYoo1liq3c/OcOyhXZUyb+0GJ7U+xbk6Gb9cMD00FnAFWxrFjHsbZGwBppaqnb/6AQ==","signature_status":"signed_v1","signed_at":"2026-07-05T09:12:29.015482Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.10858","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4f20004767d41571d9d6d9ae7879c38264b27d4c6520fa3ffc52a6e6ce9cf0e3","sha256:fadb94449e18900169cbe7c0c3c8bc56b8e48e4c8ab17c2cace98da748ee9fe1"],"state_sha256":"241e25b3c778068eea6ab2c6f6904846a5e53bb0ac83474afb9258ac77b39509"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qR4k9boj0MfQN+hQIXl9kvuWl/VY+H3s+7YulEmHA2NC//fEJlRv/YYbC5dO8awJIXXeSDTj9iTSg/o3qLTKDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T08:38:28.372312Z","bundle_sha256":"8028ec8d169f6e1c340b71f22be2ca54aa9f6fdcc0e9d8fd2d834eb896c94b6f"}}