{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:HB7NV4PM3Q3DGI2QIEPK32MXNK","short_pith_number":"pith:HB7NV4PM","canonical_record":{"source":{"id":"2510.05342","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-06T20:09:37Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"2e70e7b8b5d3a1eaadc27c6b3b54e4b75436b168500ef995c67fe57900034c60","abstract_canon_sha256":"92774fafed989d57ad664c5ddc55f6f3dea3f6abe223de8c085c32410d5a0541"},"schema_version":"1.0"},"canonical_sha256":"387edaf1ecdc36332350411eade9976abf3234fa7b908b2c41f8688f9b5dbe34","source":{"kind":"arxiv","id":"2510.05342","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2510.05342","created_at":"2026-06-02T02:04:48Z"},{"alias_kind":"arxiv_version","alias_value":"2510.05342v2","created_at":"2026-06-02T02:04:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.05342","created_at":"2026-06-02T02:04:48Z"},{"alias_kind":"pith_short_12","alias_value":"HB7NV4PM3Q3D","created_at":"2026-06-02T02:04:48Z"},{"alias_kind":"pith_short_16","alias_value":"HB7NV4PM3Q3DGI2Q","created_at":"2026-06-02T02:04:48Z"},{"alias_kind":"pith_short_8","alias_value":"HB7NV4PM","created_at":"2026-06-02T02:04:48Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:HB7NV4PM3Q3DGI2QIEPK32MXNK","target":"record","payload":{"canonical_record":{"source":{"id":"2510.05342","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-06T20:09:37Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"2e70e7b8b5d3a1eaadc27c6b3b54e4b75436b168500ef995c67fe57900034c60","abstract_canon_sha256":"92774fafed989d57ad664c5ddc55f6f3dea3f6abe223de8c085c32410d5a0541"},"schema_version":"1.0"},"canonical_sha256":"387edaf1ecdc36332350411eade9976abf3234fa7b908b2c41f8688f9b5dbe34","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T02:04:48.192422Z","signature_b64":"odH7B2kdv0F3oJ3t/iPP/g4oEYOCBwWTtxcrseO8GNO8dsU6ghqDL3Y7zelDF3uVnwbWKqS8oz5f1UcW3HUsCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"387edaf1ecdc36332350411eade9976abf3234fa7b908b2c41f8688f9b5dbe34","last_reissued_at":"2026-06-02T02:04:48.191971Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T02:04:48.191971Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2510.05342","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":"2026-06-02T02:04:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"UWQXh2uQoGWiwxshqkePORGkdU1qPBjgvFn7Cip0kvzz2P5RuokQi6drjoWl+eYDY9bGvkOCyZgMVI8/MoWOCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T07:11:46.394055Z"},"content_sha256":"1deb31aeadccaed6fedfdb1797af2f7b77bce3232dd5fd26c0e4693f690b28ce","schema_version":"1.0","event_id":"sha256:1deb31aeadccaed6fedfdb1797af2f7b77bce3232dd5fd26c0e4693f690b28ce"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:HB7NV4PM3Q3DGI2QIEPK32MXNK","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Margin Adaptive DPO: Leveraging Reward Model for Granular Control in Preference Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Hyung Gyu Rho","submitted_at":"2025-10-06T20:09:37Z","abstract_excerpt":"Direct Preference Optimization (DPO) has emerged as a simple and effective method for aligning large language models. However, its reliance on a fixed temperature parameter leads to suboptimal training on diverse preference data, causing overfitting on easy examples and under-learning from informative ones. Recent methods have emerged to counter this. While IPO addresses general overfitting, its uniform regularization can be overly conservative. The more targeted approach of $\\beta$-DPO suffers from its own limitations: its batch-level adaptation applies a single, compromised temperature to mi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.05342","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/2510.05342/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":"2026-06-02T02:04:48Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2RfSkkchOxnmsJHwguMvNGW//ogMwVbUiJUZOJVwq/AZ4nunF2D5xUrB+onJb1ITFIqiCOQ1cpdNZuP/l13DCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T07:11:46.394782Z"},"content_sha256":"fdf17aa140f5f93e2a1fa2f0c1d24c6cb1e5a80f23a5f75ffebecfe47789f11f","schema_version":"1.0","event_id":"sha256:fdf17aa140f5f93e2a1fa2f0c1d24c6cb1e5a80f23a5f75ffebecfe47789f11f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/HB7NV4PM3Q3DGI2QIEPK32MXNK/bundle.json","state_url":"https://pith.science/pith/HB7NV4PM3Q3DGI2QIEPK32MXNK/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/HB7NV4PM3Q3DGI2QIEPK32MXNK/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":"2026-06-05T07:11:46Z","links":{"resolver":"https://pith.science/pith/HB7NV4PM3Q3DGI2QIEPK32MXNK","bundle":"https://pith.science/pith/HB7NV4PM3Q3DGI2QIEPK32MXNK/bundle.json","state":"https://pith.science/pith/HB7NV4PM3Q3DGI2QIEPK32MXNK/state.json","well_known_bundle":"https://pith.science/.well-known/pith/HB7NV4PM3Q3DGI2QIEPK32MXNK/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:HB7NV4PM3Q3DGI2QIEPK32MXNK","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":"92774fafed989d57ad664c5ddc55f6f3dea3f6abe223de8c085c32410d5a0541","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-06T20:09:37Z","title_canon_sha256":"2e70e7b8b5d3a1eaadc27c6b3b54e4b75436b168500ef995c67fe57900034c60"},"schema_version":"1.0","source":{"id":"2510.05342","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2510.05342","created_at":"2026-06-02T02:04:48Z"},{"alias_kind":"arxiv_version","alias_value":"2510.05342v2","created_at":"2026-06-02T02:04:48Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.05342","created_at":"2026-06-02T02:04:48Z"},{"alias_kind":"pith_short_12","alias_value":"HB7NV4PM3Q3D","created_at":"2026-06-02T02:04:48Z"},{"alias_kind":"pith_short_16","alias_value":"HB7NV4PM3Q3DGI2Q","created_at":"2026-06-02T02:04:48Z"},{"alias_kind":"pith_short_8","alias_value":"HB7NV4PM","created_at":"2026-06-02T02:04:48Z"}],"graph_snapshots":[{"event_id":"sha256:fdf17aa140f5f93e2a1fa2f0c1d24c6cb1e5a80f23a5f75ffebecfe47789f11f","target":"graph","created_at":"2026-06-02T02:04:48Z","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/2510.05342/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Direct Preference Optimization (DPO) has emerged as a simple and effective method for aligning large language models. However, its reliance on a fixed temperature parameter leads to suboptimal training on diverse preference data, causing overfitting on easy examples and under-learning from informative ones. Recent methods have emerged to counter this. While IPO addresses general overfitting, its uniform regularization can be overly conservative. The more targeted approach of $\\beta$-DPO suffers from its own limitations: its batch-level adaptation applies a single, compromised temperature to mi","authors_text":"Hyung Gyu Rho","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-06T20:09:37Z","title":"Margin Adaptive DPO: Leveraging Reward Model for Granular Control in Preference Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.05342","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":"sha256:1deb31aeadccaed6fedfdb1797af2f7b77bce3232dd5fd26c0e4693f690b28ce","target":"record","created_at":"2026-06-02T02:04:48Z","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":"92774fafed989d57ad664c5ddc55f6f3dea3f6abe223de8c085c32410d5a0541","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-06T20:09:37Z","title_canon_sha256":"2e70e7b8b5d3a1eaadc27c6b3b54e4b75436b168500ef995c67fe57900034c60"},"schema_version":"1.0","source":{"id":"2510.05342","kind":"arxiv","version":2}},"canonical_sha256":"387edaf1ecdc36332350411eade9976abf3234fa7b908b2c41f8688f9b5dbe34","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"387edaf1ecdc36332350411eade9976abf3234fa7b908b2c41f8688f9b5dbe34","first_computed_at":"2026-06-02T02:04:48.191971Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-02T02:04:48.191971Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"odH7B2kdv0F3oJ3t/iPP/g4oEYOCBwWTtxcrseO8GNO8dsU6ghqDL3Y7zelDF3uVnwbWKqS8oz5f1UcW3HUsCw==","signature_status":"signed_v1","signed_at":"2026-06-02T02:04:48.192422Z","signed_message":"canonical_sha256_bytes"},"source_id":"2510.05342","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:1deb31aeadccaed6fedfdb1797af2f7b77bce3232dd5fd26c0e4693f690b28ce","sha256:fdf17aa140f5f93e2a1fa2f0c1d24c6cb1e5a80f23a5f75ffebecfe47789f11f"],"state_sha256":"5c369cc9a3e507253272b24a9e33e1d7b239da42f5acd9267097c0b102eb0e08"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xLJYw0qRL7P0ucOJGRqMxciKfDGU3esimGrWGxiFxZhfgmKJS83iy7H65HQ+Ji/RAWok7wFgylK9SPN+o5lVAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-05T07:11:46.398472Z","bundle_sha256":"1d008e9301a325d7b787e116f9cdb7fba14e3ee92aa7231e55f44eead7994aad"}}