{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:ON76YAMCLUOU76QH6KIMXOY43P","short_pith_number":"pith:ON76YAMC","canonical_record":{"source":{"id":"2607.25136","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-27T23:05:12Z","cross_cats_sorted":[],"title_canon_sha256":"4d0738b8acdefe97cecd9d9c5f81d2bddea5e146b58bed67498931800dbeabb3","abstract_canon_sha256":"27739a7df69ac4367fc677b42dfaac73162442f22ffce7084ce3590a0e1f8afe"},"schema_version":"1.0"},"canonical_sha256":"737fec01825d1d4ffa07f290cbbb1cdbf7ecf6d56a02af301f3792cf7351e654","source":{"kind":"arxiv","id":"2607.25136","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.25136","created_at":"2026-07-29T00:25:04Z"},{"alias_kind":"arxiv_version","alias_value":"2607.25136v1","created_at":"2026-07-29T00:25:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.25136","created_at":"2026-07-29T00:25:04Z"},{"alias_kind":"pith_short_12","alias_value":"ON76YAMCLUOU","created_at":"2026-07-29T00:25:04Z"},{"alias_kind":"pith_short_16","alias_value":"ON76YAMCLUOU76QH","created_at":"2026-07-29T00:25:04Z"},{"alias_kind":"pith_short_8","alias_value":"ON76YAMC","created_at":"2026-07-29T00:25:04Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:ON76YAMCLUOU76QH6KIMXOY43P","target":"record","payload":{"canonical_record":{"source":{"id":"2607.25136","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-27T23:05:12Z","cross_cats_sorted":[],"title_canon_sha256":"4d0738b8acdefe97cecd9d9c5f81d2bddea5e146b58bed67498931800dbeabb3","abstract_canon_sha256":"27739a7df69ac4367fc677b42dfaac73162442f22ffce7084ce3590a0e1f8afe"},"schema_version":"1.0"},"canonical_sha256":"737fec01825d1d4ffa07f290cbbb1cdbf7ecf6d56a02af301f3792cf7351e654","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-29T00:25:04.819279Z","signature_b64":"5VrnnjGu+fQ44pUUQI2tdTEnsN5ISHYyH/XeAIwMTntYwBwcEf/ir9HjfXFvQmaz35tiEX1yaURK25QBrrwYCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"737fec01825d1d4ffa07f290cbbb1cdbf7ecf6d56a02af301f3792cf7351e654","last_reissued_at":"2026-07-29T00:25:04.818406Z","signature_status":"signed_v1","first_computed_at":"2026-07-29T00:25:04.818406Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.25136","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":"2026-07-29T00:25:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8EwnO7owDU9s88reWYH8LyYAXCaHWZXZ3UNsBOyZeUdhCqpYRljZM3qjF7uG+AG2AwxkfnMm8xVB0nImGPHZCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T08:10:46.308267Z"},"content_sha256":"f2e7f3719a05a4af9676a48138da350215b2ce277b550812da094b9876643dac","schema_version":"1.0","event_id":"sha256:f2e7f3719a05a4af9676a48138da350215b2ce277b550812da094b9876643dac"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:ON76YAMCLUOU76QH6KIMXOY43P","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Less Data, Better Alignment: Data-Centric Multi-Evaluator Agreement for Preference Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chenhao Wei, Chenqian Le, Guang Yang, Haoyan Xu, Jiayi Cheng, Junhao Dong, Michael Yue, Rongchao Zhang, Runhao Li, Siheng Wang, Xupeng Chen, Yidong Wang, Yuqi Li, Zhengdao Li, Zhengtao Yao","submitted_at":"2026-07-27T23:05:12Z","abstract_excerpt":"Research on preference optimization often varies the training objective while holding the data fixed. We instead ask whether a small, high-confidence set of on-policy responses can provide a reliable learning signal. Our method, DMAPO (Data-centric Multi-evaluator Agreement for Preference Optimization), generates candidate responses from the target policy, evaluates helpfulness, factuality, and conciseness with rubric-specialized evaluators, applies a process-critic correction, and retains only high-consensus desirable or undesirable examples. This procedure accepts 1,871 of 54,236 Mistral-7B "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.25136","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/2607.25136/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":"2026-07-29T00:25:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"38KH6x2EwXRetJ3j36d/BvDpNVqqVPb7+DyuB4YETTRFa2SzIBGkRCJDfO6bNsWYLHnC7ih/m5+sa/95gj4tAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T08:10:46.309181Z"},"content_sha256":"3108452c7e1fcbc8d6390f51a2ed9ba88ea083ba6f88c425686f37a5dd03dd43","schema_version":"1.0","event_id":"sha256:3108452c7e1fcbc8d6390f51a2ed9ba88ea083ba6f88c425686f37a5dd03dd43"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ON76YAMCLUOU76QH6KIMXOY43P/bundle.json","state_url":"https://pith.science/pith/ON76YAMCLUOU76QH6KIMXOY43P/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ON76YAMCLUOU76QH6KIMXOY43P/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":"2026-08-05T08:10:46Z","links":{"resolver":"https://pith.science/pith/ON76YAMCLUOU76QH6KIMXOY43P","bundle":"https://pith.science/pith/ON76YAMCLUOU76QH6KIMXOY43P/bundle.json","state":"https://pith.science/pith/ON76YAMCLUOU76QH6KIMXOY43P/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ON76YAMCLUOU76QH6KIMXOY43P/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:ON76YAMCLUOU76QH6KIMXOY43P","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":"27739a7df69ac4367fc677b42dfaac73162442f22ffce7084ce3590a0e1f8afe","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-27T23:05:12Z","title_canon_sha256":"4d0738b8acdefe97cecd9d9c5f81d2bddea5e146b58bed67498931800dbeabb3"},"schema_version":"1.0","source":{"id":"2607.25136","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.25136","created_at":"2026-07-29T00:25:04Z"},{"alias_kind":"arxiv_version","alias_value":"2607.25136v1","created_at":"2026-07-29T00:25:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.25136","created_at":"2026-07-29T00:25:04Z"},{"alias_kind":"pith_short_12","alias_value":"ON76YAMCLUOU","created_at":"2026-07-29T00:25:04Z"},{"alias_kind":"pith_short_16","alias_value":"ON76YAMCLUOU76QH","created_at":"2026-07-29T00:25:04Z"},{"alias_kind":"pith_short_8","alias_value":"ON76YAMC","created_at":"2026-07-29T00:25:04Z"}],"graph_snapshots":[{"event_id":"sha256:3108452c7e1fcbc8d6390f51a2ed9ba88ea083ba6f88c425686f37a5dd03dd43","target":"graph","created_at":"2026-07-29T00:25:04Z","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/2607.25136/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Research on preference optimization often varies the training objective while holding the data fixed. We instead ask whether a small, high-confidence set of on-policy responses can provide a reliable learning signal. Our method, DMAPO (Data-centric Multi-evaluator Agreement for Preference Optimization), generates candidate responses from the target policy, evaluates helpfulness, factuality, and conciseness with rubric-specialized evaluators, applies a process-critic correction, and retains only high-consensus desirable or undesirable examples. This procedure accepts 1,871 of 54,236 Mistral-7B ","authors_text":"Chenhao Wei, Chenqian Le, Guang Yang, Haoyan Xu, Jiayi Cheng, Junhao Dong, Michael Yue, Rongchao Zhang, Runhao Li, Siheng Wang, Xupeng Chen, Yidong Wang, Yuqi Li, Zhengdao Li, Zhengtao Yao","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-27T23:05:12Z","title":"Less Data, Better Alignment: Data-Centric Multi-Evaluator Agreement for Preference Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.25136","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:f2e7f3719a05a4af9676a48138da350215b2ce277b550812da094b9876643dac","target":"record","created_at":"2026-07-29T00:25:04Z","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":"27739a7df69ac4367fc677b42dfaac73162442f22ffce7084ce3590a0e1f8afe","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-27T23:05:12Z","title_canon_sha256":"4d0738b8acdefe97cecd9d9c5f81d2bddea5e146b58bed67498931800dbeabb3"},"schema_version":"1.0","source":{"id":"2607.25136","kind":"arxiv","version":1}},"canonical_sha256":"737fec01825d1d4ffa07f290cbbb1cdbf7ecf6d56a02af301f3792cf7351e654","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"737fec01825d1d4ffa07f290cbbb1cdbf7ecf6d56a02af301f3792cf7351e654","first_computed_at":"2026-07-29T00:25:04.818406Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-29T00:25:04.818406Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"5VrnnjGu+fQ44pUUQI2tdTEnsN5ISHYyH/XeAIwMTntYwBwcEf/ir9HjfXFvQmaz35tiEX1yaURK25QBrrwYCg==","signature_status":"signed_v1","signed_at":"2026-07-29T00:25:04.819279Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.25136","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f2e7f3719a05a4af9676a48138da350215b2ce277b550812da094b9876643dac","sha256:3108452c7e1fcbc8d6390f51a2ed9ba88ea083ba6f88c425686f37a5dd03dd43"],"state_sha256":"8df2c6dc48fe1d78ee99f402fa1ff8963f602e72889730a23fcec66b2257dcab"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"B8nrDtm1x7BbDfJLS4sh2PJVjCd6r0aqR6c+KsxploP7ie+gMPJ8gCS3qfZtO00Qn8C63zAEWxuTU9RzVi1aCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T08:10:46.315356Z","bundle_sha256":"55d5a2820c4221541649babf7acfb2d32c5e9f3599d4d4c20554b060b22caef3"}}