{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:KDXE7KW6JUFO6DJDBLSV4SBMYO","short_pith_number":"pith:KDXE7KW6","canonical_record":{"source":{"id":"2606.19607","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-17T21:19:01Z","cross_cats_sorted":["stat.AP"],"title_canon_sha256":"58eef06d36970d9fd81cdcfdf99903d2623a991f586fc7ff8aa34493655d0ea4","abstract_canon_sha256":"04516e0854eeee81092542259dc18740c127a907fa55e910bf3c6681600088a9"},"schema_version":"1.0"},"canonical_sha256":"50ee4faade4d0aef0d230ae55e482cc3949902acebaed906184a39a571180631","source":{"kind":"arxiv","id":"2606.19607","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.19607","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"arxiv_version","alias_value":"2606.19607v1","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.19607","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"pith_short_12","alias_value":"KDXE7KW6JUFO","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"pith_short_16","alias_value":"KDXE7KW6JUFO6DJD","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"pith_short_8","alias_value":"KDXE7KW6","created_at":"2026-06-19T16:12:30Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:KDXE7KW6JUFO6DJDBLSV4SBMYO","target":"record","payload":{"canonical_record":{"source":{"id":"2606.19607","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-17T21:19:01Z","cross_cats_sorted":["stat.AP"],"title_canon_sha256":"58eef06d36970d9fd81cdcfdf99903d2623a991f586fc7ff8aa34493655d0ea4","abstract_canon_sha256":"04516e0854eeee81092542259dc18740c127a907fa55e910bf3c6681600088a9"},"schema_version":"1.0"},"canonical_sha256":"50ee4faade4d0aef0d230ae55e482cc3949902acebaed906184a39a571180631","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:12:30.053176Z","signature_b64":"hryOw/uBDQT3QUlcc9ygGXFgZdz1GA87qEV4IPqsd9i7gw0m2Iyfizca+62fuwr9tm1sGK54ZEJHjn1jMQk3Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"50ee4faade4d0aef0d230ae55e482cc3949902acebaed906184a39a571180631","last_reissued_at":"2026-06-19T16:12:30.052837Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:12:30.052837Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.19607","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":"2026-06-19T16:12:30Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"v/OJ/a4YsiV9QYtxhTqgj61xm/rmL2hOhVKtJEE4376wre3rsSdke8ViydryijIt4JnJtQ9jAatNfJ4/3YM5Bw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-04T23:11:24.489545Z"},"content_sha256":"db58d5dcb57235c38fc367d7a7d2cdddb0533707dd4d75bc8ba257f2abfcaf99","schema_version":"1.0","event_id":"sha256:db58d5dcb57235c38fc367d7a7d2cdddb0533707dd4d75bc8ba257f2abfcaf99"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:KDXE7KW6JUFO6DJDBLSV4SBMYO","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Which Pairs to Compare for LLM Post-Training?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.AP"],"primary_cat":"cs.AI","authors_text":"Jiangze Han, Vineet Goyal, Will Ma","submitted_at":"2026-06-17T21:19:01Z","abstract_excerpt":"Preference-based post-training has become a central paradigm for aligning language models. A common data-collection strategy is to generate a small set of completions for each prompt and label the resulting comparison pairs. However, human preference labels are often much more expensive than generating additional completions, suggesting a different use of the same labeling budget: generate a larger pool of completions, but label only the most informative comparison pairs. This paper studies which pairs should be compared in preference-based post-training. We formulate comparison curation as a "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.19607","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/2606.19607/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":"2026-06-19T16:12:30Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2j6zrfh5jkXtizwWc0wdOphlEPK+i7QyiK9TAU47HuXvAFAA0ab15XRNQyZUflmOlD9NY0GGT10P3zE+f6BLDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-04T23:11:24.489915Z"},"content_sha256":"ae5344a903b5460c883d1b34e83fe303b2ebed1430555fe09466e8afcd1c0009","schema_version":"1.0","event_id":"sha256:ae5344a903b5460c883d1b34e83fe303b2ebed1430555fe09466e8afcd1c0009"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/KDXE7KW6JUFO6DJDBLSV4SBMYO/bundle.json","state_url":"https://pith.science/pith/KDXE7KW6JUFO6DJDBLSV4SBMYO/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/KDXE7KW6JUFO6DJDBLSV4SBMYO/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":"2026-07-04T23:11:24Z","links":{"resolver":"https://pith.science/pith/KDXE7KW6JUFO6DJDBLSV4SBMYO","bundle":"https://pith.science/pith/KDXE7KW6JUFO6DJDBLSV4SBMYO/bundle.json","state":"https://pith.science/pith/KDXE7KW6JUFO6DJDBLSV4SBMYO/state.json","well_known_bundle":"https://pith.science/.well-known/pith/KDXE7KW6JUFO6DJDBLSV4SBMYO/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:KDXE7KW6JUFO6DJDBLSV4SBMYO","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":"04516e0854eeee81092542259dc18740c127a907fa55e910bf3c6681600088a9","cross_cats_sorted":["stat.AP"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-17T21:19:01Z","title_canon_sha256":"58eef06d36970d9fd81cdcfdf99903d2623a991f586fc7ff8aa34493655d0ea4"},"schema_version":"1.0","source":{"id":"2606.19607","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.19607","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"arxiv_version","alias_value":"2606.19607v1","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.19607","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"pith_short_12","alias_value":"KDXE7KW6JUFO","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"pith_short_16","alias_value":"KDXE7KW6JUFO6DJD","created_at":"2026-06-19T16:12:30Z"},{"alias_kind":"pith_short_8","alias_value":"KDXE7KW6","created_at":"2026-06-19T16:12:30Z"}],"graph_snapshots":[{"event_id":"sha256:ae5344a903b5460c883d1b34e83fe303b2ebed1430555fe09466e8afcd1c0009","target":"graph","created_at":"2026-06-19T16:12:30Z","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/2606.19607/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Preference-based post-training has become a central paradigm for aligning language models. A common data-collection strategy is to generate a small set of completions for each prompt and label the resulting comparison pairs. However, human preference labels are often much more expensive than generating additional completions, suggesting a different use of the same labeling budget: generate a larger pool of completions, but label only the most informative comparison pairs. This paper studies which pairs should be compared in preference-based post-training. We formulate comparison curation as a ","authors_text":"Jiangze Han, Vineet Goyal, Will Ma","cross_cats":["stat.AP"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-17T21:19:01Z","title":"Which Pairs to Compare for LLM Post-Training?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.19607","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:db58d5dcb57235c38fc367d7a7d2cdddb0533707dd4d75bc8ba257f2abfcaf99","target":"record","created_at":"2026-06-19T16:12:30Z","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":"04516e0854eeee81092542259dc18740c127a907fa55e910bf3c6681600088a9","cross_cats_sorted":["stat.AP"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-17T21:19:01Z","title_canon_sha256":"58eef06d36970d9fd81cdcfdf99903d2623a991f586fc7ff8aa34493655d0ea4"},"schema_version":"1.0","source":{"id":"2606.19607","kind":"arxiv","version":1}},"canonical_sha256":"50ee4faade4d0aef0d230ae55e482cc3949902acebaed906184a39a571180631","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"50ee4faade4d0aef0d230ae55e482cc3949902acebaed906184a39a571180631","first_computed_at":"2026-06-19T16:12:30.052837Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-19T16:12:30.052837Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hryOw/uBDQT3QUlcc9ygGXFgZdz1GA87qEV4IPqsd9i7gw0m2Iyfizca+62fuwr9tm1sGK54ZEJHjn1jMQk3Cw==","signature_status":"signed_v1","signed_at":"2026-06-19T16:12:30.053176Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.19607","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:db58d5dcb57235c38fc367d7a7d2cdddb0533707dd4d75bc8ba257f2abfcaf99","sha256:ae5344a903b5460c883d1b34e83fe303b2ebed1430555fe09466e8afcd1c0009"],"state_sha256":"acd90ccb1337ef94d5f51ff61f70a9f490c1fdf898946bce526605934eb836b0"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"UhFQcsw+d4s/NzfCc28NZRPoev3HAUJ/hwL3+fnvY+HMBDtd6IOvvc59cjS0bUhfd+SIcB0cjAwO8Hvfa7P9CA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-04T23:11:24.491849Z","bundle_sha256":"21c8c37a4aa2ec4ecc538e001fe68652f92da1b8b01fc9e644ab06fafb6bcccf"}}