{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:J37KW4IQF25KHSSSFQONMAMNID","short_pith_number":"pith:J37KW4IQ","canonical_record":{"source":{"id":"2406.07295","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-11T14:24:00Z","cross_cats_sorted":[],"title_canon_sha256":"57abf7b6a6847f0b7c18cceb6173f72d4434276742725eb335bac8fcaf0ccf6d","abstract_canon_sha256":"7bb4e47ab64e8e5393e9239ecf4fa1ce61e861dd66687b21c0ba42120d8ac482"},"schema_version":"1.0"},"canonical_sha256":"4efeab71102ebaa3ca522c1cd6018d40ec58533bf6825e9a78b5d74e4f59aa66","source":{"kind":"arxiv","id":"2406.07295","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.07295","created_at":"2026-07-05T08:30:52Z"},{"alias_kind":"arxiv_version","alias_value":"2406.07295v2","created_at":"2026-07-05T08:30:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.07295","created_at":"2026-07-05T08:30:52Z"},{"alias_kind":"pith_short_12","alias_value":"J37KW4IQF25K","created_at":"2026-07-05T08:30:52Z"},{"alias_kind":"pith_short_16","alias_value":"J37KW4IQF25KHSSS","created_at":"2026-07-05T08:30:52Z"},{"alias_kind":"pith_short_8","alias_value":"J37KW4IQ","created_at":"2026-07-05T08:30:52Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:J37KW4IQF25KHSSSFQONMAMNID","target":"record","payload":{"canonical_record":{"source":{"id":"2406.07295","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-11T14:24:00Z","cross_cats_sorted":[],"title_canon_sha256":"57abf7b6a6847f0b7c18cceb6173f72d4434276742725eb335bac8fcaf0ccf6d","abstract_canon_sha256":"7bb4e47ab64e8e5393e9239ecf4fa1ce61e861dd66687b21c0ba42120d8ac482"},"schema_version":"1.0"},"canonical_sha256":"4efeab71102ebaa3ca522c1cd6018d40ec58533bf6825e9a78b5d74e4f59aa66","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:30:52.706467Z","signature_b64":"pCm9qGK7eI7vpvcUaHKnh8+wV2GL+5KW7ASgdMSZ0cVa49HK3ZvpcrHclO6HN2B0UymLEIdjyS/oGS2m17EOCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4efeab71102ebaa3ca522c1cd6018d40ec58533bf6825e9a78b5d74e4f59aa66","last_reissued_at":"2026-07-05T08:30:52.705999Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:30:52.705999Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.07295","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":"2026-07-05T08:30:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"o/ulr5DxSO6Xqrc7tLCMTkVdonFoUKpgp6wQ4O4bdW36cQVVvZHG1szv+UOplD8yu3JcqNXv9xuTMe9zPL8IBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T08:18:26.815581Z"},"content_sha256":"0cfc0d9cc0f902a01859ed199e386e34b9f895f342a2342648a81ea3b2386b48","schema_version":"1.0","event_id":"sha256:0cfc0d9cc0f902a01859ed199e386e34b9f895f342a2342648a81ea3b2386b48"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:J37KW4IQF25KHSSSFQONMAMNID","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Multi-objective Reinforcement learning from AI Feedback","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Marcus Williams","submitted_at":"2024-06-11T14:24:00Z","abstract_excerpt":"This paper presents Multi-Objective Reinforcement Learning from AI Feedback (MORLAIF), a novel approach to improving the alignment and performance of language models trained using reinforcement learning from AI feedback (RLAIF). In contrast to standard approaches that train a single preference model to represent all human preferences, MORLAIF decomposes this task into multiple simpler principles, such as toxicity, factuality, and sycophancy. Separate preference models are trained for each principle using feedback from GPT-3.5-Turbo. These preference model scores are then combined using differe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.07295","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/2406.07295/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":"2026-07-05T08:30:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xOA5Eab+7+W8LQ+BM7+kebLt8yA8yCzyw+OvJ0xrlwVUrn1LEOYtodc46zy4G6Jj7WneN6DoJE06ADPBBKLXAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T08:18:26.816624Z"},"content_sha256":"c128a6d6473281d37aa1807609088270689df15a630b739ac15f1e6e3858d4a2","schema_version":"1.0","event_id":"sha256:c128a6d6473281d37aa1807609088270689df15a630b739ac15f1e6e3858d4a2"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/J37KW4IQF25KHSSSFQONMAMNID/bundle.json","state_url":"https://pith.science/pith/J37KW4IQF25KHSSSFQONMAMNID/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/J37KW4IQF25KHSSSFQONMAMNID/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":"2026-08-10T08:18:26Z","links":{"resolver":"https://pith.science/pith/J37KW4IQF25KHSSSFQONMAMNID","bundle":"https://pith.science/pith/J37KW4IQF25KHSSSFQONMAMNID/bundle.json","state":"https://pith.science/pith/J37KW4IQF25KHSSSFQONMAMNID/state.json","well_known_bundle":"https://pith.science/.well-known/pith/J37KW4IQF25KHSSSFQONMAMNID/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:J37KW4IQF25KHSSSFQONMAMNID","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":"7bb4e47ab64e8e5393e9239ecf4fa1ce61e861dd66687b21c0ba42120d8ac482","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-11T14:24:00Z","title_canon_sha256":"57abf7b6a6847f0b7c18cceb6173f72d4434276742725eb335bac8fcaf0ccf6d"},"schema_version":"1.0","source":{"id":"2406.07295","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.07295","created_at":"2026-07-05T08:30:52Z"},{"alias_kind":"arxiv_version","alias_value":"2406.07295v2","created_at":"2026-07-05T08:30:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.07295","created_at":"2026-07-05T08:30:52Z"},{"alias_kind":"pith_short_12","alias_value":"J37KW4IQF25K","created_at":"2026-07-05T08:30:52Z"},{"alias_kind":"pith_short_16","alias_value":"J37KW4IQF25KHSSS","created_at":"2026-07-05T08:30:52Z"},{"alias_kind":"pith_short_8","alias_value":"J37KW4IQ","created_at":"2026-07-05T08:30:52Z"}],"graph_snapshots":[{"event_id":"sha256:c128a6d6473281d37aa1807609088270689df15a630b739ac15f1e6e3858d4a2","target":"graph","created_at":"2026-07-05T08:30:52Z","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/2406.07295/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper presents Multi-Objective Reinforcement Learning from AI Feedback (MORLAIF), a novel approach to improving the alignment and performance of language models trained using reinforcement learning from AI feedback (RLAIF). In contrast to standard approaches that train a single preference model to represent all human preferences, MORLAIF decomposes this task into multiple simpler principles, such as toxicity, factuality, and sycophancy. Separate preference models are trained for each principle using feedback from GPT-3.5-Turbo. These preference model scores are then combined using differe","authors_text":"Marcus Williams","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-11T14:24:00Z","title":"Multi-objective Reinforcement learning from AI Feedback"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.07295","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":"sha256:0cfc0d9cc0f902a01859ed199e386e34b9f895f342a2342648a81ea3b2386b48","target":"record","created_at":"2026-07-05T08:30:52Z","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":"7bb4e47ab64e8e5393e9239ecf4fa1ce61e861dd66687b21c0ba42120d8ac482","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-06-11T14:24:00Z","title_canon_sha256":"57abf7b6a6847f0b7c18cceb6173f72d4434276742725eb335bac8fcaf0ccf6d"},"schema_version":"1.0","source":{"id":"2406.07295","kind":"arxiv","version":2}},"canonical_sha256":"4efeab71102ebaa3ca522c1cd6018d40ec58533bf6825e9a78b5d74e4f59aa66","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4efeab71102ebaa3ca522c1cd6018d40ec58533bf6825e9a78b5d74e4f59aa66","first_computed_at":"2026-07-05T08:30:52.705999Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:30:52.705999Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"pCm9qGK7eI7vpvcUaHKnh8+wV2GL+5KW7ASgdMSZ0cVa49HK3ZvpcrHclO6HN2B0UymLEIdjyS/oGS2m17EOCw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:30:52.706467Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.07295","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0cfc0d9cc0f902a01859ed199e386e34b9f895f342a2342648a81ea3b2386b48","sha256:c128a6d6473281d37aa1807609088270689df15a630b739ac15f1e6e3858d4a2"],"state_sha256":"c51dda82657c0478d36992da65dc7bf6a16352c0304cd3b97cfbf9e682e41288"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"RPcdGE3/2ORhYWj2SNIEED+qqLtA//wYVaNO9eU9sNLB7xnyF+Ca1V5kVSnJAQrjk0am0eDXH3jMLcQ6vUQYDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T08:18:26.827004Z","bundle_sha256":"101cc014085cdc1cdd715b65b23dc3c675dc8dc15fdb2f55513f0229d0cdf41b"}}