{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:EZ2XCRJSIBLT5CVEWKQEJ2BWBP","short_pith_number":"pith:EZ2XCRJS","canonical_record":{"source":{"id":"2506.15706","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-30T08:42:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5ceffa782a9e855d8b577d0af99e06bcde2830ad01560556a82315dbb59d819a","abstract_canon_sha256":"561a87d10060d341195318a94dcc4e8907b521a3a9aa630c1a6639a941a509ed"},"schema_version":"1.0"},"canonical_sha256":"267571453240573e8aa4b2a044e8360bf61f4594d8648958205bb3d9c17f1ab0","source":{"kind":"arxiv","id":"2506.15706","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.15706","created_at":"2026-07-05T11:23:50Z"},{"alias_kind":"arxiv_version","alias_value":"2506.15706v1","created_at":"2026-07-05T11:23:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.15706","created_at":"2026-07-05T11:23:50Z"},{"alias_kind":"pith_short_12","alias_value":"EZ2XCRJSIBLT","created_at":"2026-07-05T11:23:50Z"},{"alias_kind":"pith_short_16","alias_value":"EZ2XCRJSIBLT5CVE","created_at":"2026-07-05T11:23:50Z"},{"alias_kind":"pith_short_8","alias_value":"EZ2XCRJS","created_at":"2026-07-05T11:23:50Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:EZ2XCRJSIBLT5CVEWKQEJ2BWBP","target":"record","payload":{"canonical_record":{"source":{"id":"2506.15706","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-30T08:42:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5ceffa782a9e855d8b577d0af99e06bcde2830ad01560556a82315dbb59d819a","abstract_canon_sha256":"561a87d10060d341195318a94dcc4e8907b521a3a9aa630c1a6639a941a509ed"},"schema_version":"1.0"},"canonical_sha256":"267571453240573e8aa4b2a044e8360bf61f4594d8648958205bb3d9c17f1ab0","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:23:50.737541Z","signature_b64":"hWr0foNHIsSgv5kDR4nB0q3yQgDhY6cgAM1EQ6pHjksbex0tYzJsz7NYPsOanuC2nfaVZxQKCEO2w+vgtTSODw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"267571453240573e8aa4b2a044e8360bf61f4594d8648958205bb3d9c17f1ab0","last_reissued_at":"2026-07-05T11:23:50.737055Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:23:50.737055Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2506.15706","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":"2026-07-05T11:23:50Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bzjiLqB8q9A0Bxnckk0YfiINur5vGPIefSK3CsXx+wECP1MZ48JBYTWQDRDcZ5dG27c53fNROCTjLBwQmiRiDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T04:15:13.874575Z"},"content_sha256":"ff292ee434f233d3676a6b4e92686797646063903001cb144cbd8c0158d673d8","schema_version":"1.0","event_id":"sha256:ff292ee434f233d3676a6b4e92686797646063903001cb144cbd8c0158d673d8"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:EZ2XCRJSIBLT5CVEWKQEJ2BWBP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"MDPO: Multi-Granularity Direct Preference Optimization for Mathematical Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Yunze Lin","submitted_at":"2025-05-30T08:42:14Z","abstract_excerpt":"Mathematical reasoning presents a significant challenge for Large Language Models (LLMs) as it requires ensuring the correctness of each reasoning step. Researchers have been strengthening the mathematical reasoning abilities of LLMs through supervised fine-tuning, but due to the inability to suppress incorrect outputs, illusions can easily arise. Recently, Direct Preference Optimization (DPO) has been widely adopted for aligning human intent by using preference data to prevent LLMs from generating incorrect outputs. However, it has shown limited benefits in long-chain mathematical reasoning, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.15706","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/2506.15706/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":"2026-07-05T11:23:50Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"+V+s0eHW9XxdGpGYr4PxC3Np9D+0fRASOXydb9oBNM9uzANOdqJzo0qNlex2skbnz6DjdPIm125GxT7zfxjHDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T04:15:13.875072Z"},"content_sha256":"d6a253f52645283ad2db3d10502c353c65d7f94795efb8cde29930facb7fe05d","schema_version":"1.0","event_id":"sha256:d6a253f52645283ad2db3d10502c353c65d7f94795efb8cde29930facb7fe05d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/EZ2XCRJSIBLT5CVEWKQEJ2BWBP/bundle.json","state_url":"https://pith.science/pith/EZ2XCRJSIBLT5CVEWKQEJ2BWBP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/EZ2XCRJSIBLT5CVEWKQEJ2BWBP/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":"2026-08-09T04:15:13Z","links":{"resolver":"https://pith.science/pith/EZ2XCRJSIBLT5CVEWKQEJ2BWBP","bundle":"https://pith.science/pith/EZ2XCRJSIBLT5CVEWKQEJ2BWBP/bundle.json","state":"https://pith.science/pith/EZ2XCRJSIBLT5CVEWKQEJ2BWBP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/EZ2XCRJSIBLT5CVEWKQEJ2BWBP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:EZ2XCRJSIBLT5CVEWKQEJ2BWBP","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":"561a87d10060d341195318a94dcc4e8907b521a3a9aa630c1a6639a941a509ed","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-30T08:42:14Z","title_canon_sha256":"5ceffa782a9e855d8b577d0af99e06bcde2830ad01560556a82315dbb59d819a"},"schema_version":"1.0","source":{"id":"2506.15706","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.15706","created_at":"2026-07-05T11:23:50Z"},{"alias_kind":"arxiv_version","alias_value":"2506.15706v1","created_at":"2026-07-05T11:23:50Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.15706","created_at":"2026-07-05T11:23:50Z"},{"alias_kind":"pith_short_12","alias_value":"EZ2XCRJSIBLT","created_at":"2026-07-05T11:23:50Z"},{"alias_kind":"pith_short_16","alias_value":"EZ2XCRJSIBLT5CVE","created_at":"2026-07-05T11:23:50Z"},{"alias_kind":"pith_short_8","alias_value":"EZ2XCRJS","created_at":"2026-07-05T11:23:50Z"}],"graph_snapshots":[{"event_id":"sha256:d6a253f52645283ad2db3d10502c353c65d7f94795efb8cde29930facb7fe05d","target":"graph","created_at":"2026-07-05T11:23:50Z","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/2506.15706/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Mathematical reasoning presents a significant challenge for Large Language Models (LLMs) as it requires ensuring the correctness of each reasoning step. Researchers have been strengthening the mathematical reasoning abilities of LLMs through supervised fine-tuning, but due to the inability to suppress incorrect outputs, illusions can easily arise. Recently, Direct Preference Optimization (DPO) has been widely adopted for aligning human intent by using preference data to prevent LLMs from generating incorrect outputs. However, it has shown limited benefits in long-chain mathematical reasoning, ","authors_text":"Yunze Lin","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-30T08:42:14Z","title":"MDPO: Multi-Granularity Direct Preference Optimization for Mathematical Reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.15706","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:ff292ee434f233d3676a6b4e92686797646063903001cb144cbd8c0158d673d8","target":"record","created_at":"2026-07-05T11:23:50Z","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":"561a87d10060d341195318a94dcc4e8907b521a3a9aa630c1a6639a941a509ed","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-05-30T08:42:14Z","title_canon_sha256":"5ceffa782a9e855d8b577d0af99e06bcde2830ad01560556a82315dbb59d819a"},"schema_version":"1.0","source":{"id":"2506.15706","kind":"arxiv","version":1}},"canonical_sha256":"267571453240573e8aa4b2a044e8360bf61f4594d8648958205bb3d9c17f1ab0","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"267571453240573e8aa4b2a044e8360bf61f4594d8648958205bb3d9c17f1ab0","first_computed_at":"2026-07-05T11:23:50.737055Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:23:50.737055Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hWr0foNHIsSgv5kDR4nB0q3yQgDhY6cgAM1EQ6pHjksbex0tYzJsz7NYPsOanuC2nfaVZxQKCEO2w+vgtTSODw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:23:50.737541Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.15706","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ff292ee434f233d3676a6b4e92686797646063903001cb144cbd8c0158d673d8","sha256:d6a253f52645283ad2db3d10502c353c65d7f94795efb8cde29930facb7fe05d"],"state_sha256":"713dbba073abcd52c401eb307729db01cbf94c9e92a75a442b4afea646e0cabf"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"C6qdRvWCC1QKOCV1pVWKgZicy/b/lk91xAPC1YGhcHNB3LhElbFhmoyXJvOQqdpvxTPRDg4nPMaDtQtzD7S/Bg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T04:15:13.879103Z","bundle_sha256":"587374fef38731f73d5ddbc282827013f52e86473c887510164557635786a256"}}