{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:XCEVIVJQCYY7QGYRWLECBZIZHL","short_pith_number":"pith:XCEVIVJQ","canonical_record":{"source":{"id":"2407.16008","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-22T19:21:55Z","cross_cats_sorted":[],"title_canon_sha256":"c25902bc2fa369a3e310c78c42c8cf59f7b5e7f1cb57eae199dc274a90081efd","abstract_canon_sha256":"1d29e5aacc5e0f2a96cf1d803f3ec908ecef8e61b0a105306757a3279c11b3b4"},"schema_version":"1.0"},"canonical_sha256":"b8895455301631f81b11b2c820e5193afe5f6fc39075cc0b937e2b698a876d27","source":{"kind":"arxiv","id":"2407.16008","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.16008","created_at":"2026-07-05T10:31:45Z"},{"alias_kind":"arxiv_version","alias_value":"2407.16008v2","created_at":"2026-07-05T10:31:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.16008","created_at":"2026-07-05T10:31:45Z"},{"alias_kind":"pith_short_12","alias_value":"XCEVIVJQCYY7","created_at":"2026-07-05T10:31:45Z"},{"alias_kind":"pith_short_16","alias_value":"XCEVIVJQCYY7QGYR","created_at":"2026-07-05T10:31:45Z"},{"alias_kind":"pith_short_8","alias_value":"XCEVIVJQ","created_at":"2026-07-05T10:31:45Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:XCEVIVJQCYY7QGYRWLECBZIZHL","target":"record","payload":{"canonical_record":{"source":{"id":"2407.16008","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-22T19:21:55Z","cross_cats_sorted":[],"title_canon_sha256":"c25902bc2fa369a3e310c78c42c8cf59f7b5e7f1cb57eae199dc274a90081efd","abstract_canon_sha256":"1d29e5aacc5e0f2a96cf1d803f3ec908ecef8e61b0a105306757a3279c11b3b4"},"schema_version":"1.0"},"canonical_sha256":"b8895455301631f81b11b2c820e5193afe5f6fc39075cc0b937e2b698a876d27","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:31:45.363576Z","signature_b64":"o600cEFW1UdfAypX6YdBF0DCHk/wfJI9q+UHWC3Yf4pc3176UKkHWRPe8hsdcqX6c9cGtbdAlr3Erp6P+YWTCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b8895455301631f81b11b2c820e5193afe5f6fc39075cc0b937e2b698a876d27","last_reissued_at":"2026-07-05T10:31:45.362743Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:31:45.362743Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2407.16008","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":"2026-07-05T10:31:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"umkOiqZjHCiMkRulcaVdHsJUjtbLvwcHNibiRY2YdrG/UAwRbSP2qNWSXLkWBPilmujlJ3h4S2AQMIOXBU//Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T15:06:53.509825Z"},"content_sha256":"12eb5ccb812116a37b41c825a1685179886030b51d01fcc05482d05fcc20104d","schema_version":"1.0","event_id":"sha256:12eb5ccb812116a37b41c825a1685179886030b51d01fcc05482d05fcc20104d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:XCEVIVJQCYY7QGYRWLECBZIZHL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Boosting Reward Model with Preference-Conditional Multi-Aspect Synthetic Data Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Carl Yang, Jiaming Shen, Michael Bendersky, Ran Xu, Simon Baumgartner, Tianqi Liu, Yennie Jun, Yi Liang, Zhen Qin","submitted_at":"2024-07-22T19:21:55Z","abstract_excerpt":"Reward models (RMs) are crucial for aligning large language models (LLMs) with human preferences. They are trained using preference datasets where each example consists of one input prompt, two responses, and a preference label. As curating a high-quality human labeled preference dataset is both time-consuming and expensive, people often rely on existing powerful LLMs for preference label generation. This can potentially introduce noise and impede RM training. In this work, we present RMBoost, a novel synthetic preference data generation paradigm to boost reward model quality. Unlike tradition"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.16008","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/2407.16008/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":"2026-07-05T10:31:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mVj5UUp3gElPmNwknSySXTxKQsCgrmF0HQH9kq2STlxEEBSGX+OReIz8CC7NKa9pcRzojCt4oflhh5tzF0HRBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T15:06:53.510197Z"},"content_sha256":"794656dca94ebe8de1ba23e4294e6c9202c9cf7fa8fe48decb07896379a05a66","schema_version":"1.0","event_id":"sha256:794656dca94ebe8de1ba23e4294e6c9202c9cf7fa8fe48decb07896379a05a66"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XCEVIVJQCYY7QGYRWLECBZIZHL/bundle.json","state_url":"https://pith.science/pith/XCEVIVJQCYY7QGYRWLECBZIZHL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XCEVIVJQCYY7QGYRWLECBZIZHL/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":"2026-07-06T15:06:53Z","links":{"resolver":"https://pith.science/pith/XCEVIVJQCYY7QGYRWLECBZIZHL","bundle":"https://pith.science/pith/XCEVIVJQCYY7QGYRWLECBZIZHL/bundle.json","state":"https://pith.science/pith/XCEVIVJQCYY7QGYRWLECBZIZHL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XCEVIVJQCYY7QGYRWLECBZIZHL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:XCEVIVJQCYY7QGYRWLECBZIZHL","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":"1d29e5aacc5e0f2a96cf1d803f3ec908ecef8e61b0a105306757a3279c11b3b4","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-22T19:21:55Z","title_canon_sha256":"c25902bc2fa369a3e310c78c42c8cf59f7b5e7f1cb57eae199dc274a90081efd"},"schema_version":"1.0","source":{"id":"2407.16008","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.16008","created_at":"2026-07-05T10:31:45Z"},{"alias_kind":"arxiv_version","alias_value":"2407.16008v2","created_at":"2026-07-05T10:31:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.16008","created_at":"2026-07-05T10:31:45Z"},{"alias_kind":"pith_short_12","alias_value":"XCEVIVJQCYY7","created_at":"2026-07-05T10:31:45Z"},{"alias_kind":"pith_short_16","alias_value":"XCEVIVJQCYY7QGYR","created_at":"2026-07-05T10:31:45Z"},{"alias_kind":"pith_short_8","alias_value":"XCEVIVJQ","created_at":"2026-07-05T10:31:45Z"}],"graph_snapshots":[{"event_id":"sha256:794656dca94ebe8de1ba23e4294e6c9202c9cf7fa8fe48decb07896379a05a66","target":"graph","created_at":"2026-07-05T10:31:45Z","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/2407.16008/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reward models (RMs) are crucial for aligning large language models (LLMs) with human preferences. They are trained using preference datasets where each example consists of one input prompt, two responses, and a preference label. As curating a high-quality human labeled preference dataset is both time-consuming and expensive, people often rely on existing powerful LLMs for preference label generation. This can potentially introduce noise and impede RM training. In this work, we present RMBoost, a novel synthetic preference data generation paradigm to boost reward model quality. Unlike tradition","authors_text":"Carl Yang, Jiaming Shen, Michael Bendersky, Ran Xu, Simon Baumgartner, Tianqi Liu, Yennie Jun, Yi Liang, Zhen Qin","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-22T19:21:55Z","title":"Boosting Reward Model with Preference-Conditional Multi-Aspect Synthetic Data Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.16008","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":"sha256:12eb5ccb812116a37b41c825a1685179886030b51d01fcc05482d05fcc20104d","target":"record","created_at":"2026-07-05T10:31:45Z","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":"1d29e5aacc5e0f2a96cf1d803f3ec908ecef8e61b0a105306757a3279c11b3b4","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-07-22T19:21:55Z","title_canon_sha256":"c25902bc2fa369a3e310c78c42c8cf59f7b5e7f1cb57eae199dc274a90081efd"},"schema_version":"1.0","source":{"id":"2407.16008","kind":"arxiv","version":2}},"canonical_sha256":"b8895455301631f81b11b2c820e5193afe5f6fc39075cc0b937e2b698a876d27","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b8895455301631f81b11b2c820e5193afe5f6fc39075cc0b937e2b698a876d27","first_computed_at":"2026-07-05T10:31:45.362743Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:31:45.362743Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"o600cEFW1UdfAypX6YdBF0DCHk/wfJI9q+UHWC3Yf4pc3176UKkHWRPe8hsdcqX6c9cGtbdAlr3Erp6P+YWTCw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:31:45.363576Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.16008","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:12eb5ccb812116a37b41c825a1685179886030b51d01fcc05482d05fcc20104d","sha256:794656dca94ebe8de1ba23e4294e6c9202c9cf7fa8fe48decb07896379a05a66"],"state_sha256":"05dd481c912c9416600c770de191d448677b2faacb7a626d40bdb3f7fb86d829"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tFl9u+j+QyxFJhvEaGkA4g5ulYMAdCb9WCxxBcByzqKLgZmaWQOAMEqdkJ926Ft8di9i5wzJLPuQnwi4Z9DMBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T15:06:53.512243Z","bundle_sha256":"c74546d86a5fc9df3e3576dbd452e934c19aae077606ae4e6e8fabe0393dd11e"}}