{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:EFRLLTVKH6RWVDJTVT2YZZ3RNR","short_pith_number":"pith:EFRLLTVK","canonical_record":{"source":{"id":"2412.03822","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-05T02:35:46Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6fef43bb798a0b30b4fee70c7b608c80a5136d052a4314acde8d2552edcb9ae8","abstract_canon_sha256":"73d2f87e9ec1c17fa4a766c648162869ef068b0f5076e82f932ffeecf2464ad7"},"schema_version":"1.0"},"canonical_sha256":"2162b5ceaa3fa36a8d33acf58ce7716c71edd82cde0a6a4022e79c99c2a89020","source":{"kind":"arxiv","id":"2412.03822","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.03822","created_at":"2026-07-05T09:44:41Z"},{"alias_kind":"arxiv_version","alias_value":"2412.03822v1","created_at":"2026-07-05T09:44:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.03822","created_at":"2026-07-05T09:44:41Z"},{"alias_kind":"pith_short_12","alias_value":"EFRLLTVKH6RW","created_at":"2026-07-05T09:44:41Z"},{"alias_kind":"pith_short_16","alias_value":"EFRLLTVKH6RWVDJT","created_at":"2026-07-05T09:44:41Z"},{"alias_kind":"pith_short_8","alias_value":"EFRLLTVK","created_at":"2026-07-05T09:44:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:EFRLLTVKH6RWVDJTVT2YZZ3RNR","target":"record","payload":{"canonical_record":{"source":{"id":"2412.03822","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-05T02:35:46Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6fef43bb798a0b30b4fee70c7b608c80a5136d052a4314acde8d2552edcb9ae8","abstract_canon_sha256":"73d2f87e9ec1c17fa4a766c648162869ef068b0f5076e82f932ffeecf2464ad7"},"schema_version":"1.0"},"canonical_sha256":"2162b5ceaa3fa36a8d33acf58ce7716c71edd82cde0a6a4022e79c99c2a89020","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:44:41.780005Z","signature_b64":"H9MausHrtD7dyvFrFh71eAw+vujF8c08xoKBdTtDGNVOguzrC1WOEaoqGbSmIME4Wpwt7530w0VFtdoJd6zrAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2162b5ceaa3fa36a8d33acf58ce7716c71edd82cde0a6a4022e79c99c2a89020","last_reissued_at":"2026-07-05T09:44:41.779539Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:44:41.779539Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.03822","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":"2026-07-05T09:44:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uFVAoSFnXo59tCEGfDDa8isJ01gk+90vVnKCrZ9m89++ChP1849xrhmoSvSNeRIBd0OY68AqD9hVWGBgboJBDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-12T22:46:19.835039Z"},"content_sha256":"7c98fc5f03194eb0e3289f53147d42ff07efbb7ec1b425da175d77488699984a","schema_version":"1.0","event_id":"sha256:7c98fc5f03194eb0e3289f53147d42ff07efbb7ec1b425da175d77488699984a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:EFRLLTVKH6RWVDJTVT2YZZ3RNR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Beyond the Binary: Capturing Diverse Preferences With Reward Regularization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chuanyang Jin, Hannah Rose Kirk, He He, Vishakh Padmakumar","submitted_at":"2024-12-05T02:35:46Z","abstract_excerpt":"Large language models (LLMs) are increasingly deployed via public-facing interfaces to interact with millions of users, each with diverse preferences. Despite this, preference tuning of LLMs predominantly relies on reward models trained using binary judgments where annotators select the preferred choice out of pairs of model outputs. In this work, we argue that this reliance on binary choices does not capture the broader, aggregate preferences of the target user in real-world tasks. We propose a taxonomy that identifies two dimensions of subjectivity where different users disagree on the prefe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.03822","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/2412.03822/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":"2026-07-05T09:44:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qQ3g+8mFxqr7sF+qtP/yBMgc2doznR43tRxfaSx3OYMTDHWhYI7kGWgmnFFnfGwalWP5j1+jenI9BXKMhs25Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-12T22:46:19.835423Z"},"content_sha256":"4fe41596bbd2c465c231b2badd4d10bceb0bd0e24efe5a3083ef7e37d556c0ef","schema_version":"1.0","event_id":"sha256:4fe41596bbd2c465c231b2badd4d10bceb0bd0e24efe5a3083ef7e37d556c0ef"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/EFRLLTVKH6RWVDJTVT2YZZ3RNR/bundle.json","state_url":"https://pith.science/pith/EFRLLTVKH6RWVDJTVT2YZZ3RNR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/EFRLLTVKH6RWVDJTVT2YZZ3RNR/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":"2026-08-12T22:46:19Z","links":{"resolver":"https://pith.science/pith/EFRLLTVKH6RWVDJTVT2YZZ3RNR","bundle":"https://pith.science/pith/EFRLLTVKH6RWVDJTVT2YZZ3RNR/bundle.json","state":"https://pith.science/pith/EFRLLTVKH6RWVDJTVT2YZZ3RNR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/EFRLLTVKH6RWVDJTVT2YZZ3RNR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:EFRLLTVKH6RWVDJTVT2YZZ3RNR","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":"73d2f87e9ec1c17fa4a766c648162869ef068b0f5076e82f932ffeecf2464ad7","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-05T02:35:46Z","title_canon_sha256":"6fef43bb798a0b30b4fee70c7b608c80a5136d052a4314acde8d2552edcb9ae8"},"schema_version":"1.0","source":{"id":"2412.03822","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.03822","created_at":"2026-07-05T09:44:41Z"},{"alias_kind":"arxiv_version","alias_value":"2412.03822v1","created_at":"2026-07-05T09:44:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.03822","created_at":"2026-07-05T09:44:41Z"},{"alias_kind":"pith_short_12","alias_value":"EFRLLTVKH6RW","created_at":"2026-07-05T09:44:41Z"},{"alias_kind":"pith_short_16","alias_value":"EFRLLTVKH6RWVDJT","created_at":"2026-07-05T09:44:41Z"},{"alias_kind":"pith_short_8","alias_value":"EFRLLTVK","created_at":"2026-07-05T09:44:41Z"}],"graph_snapshots":[{"event_id":"sha256:4fe41596bbd2c465c231b2badd4d10bceb0bd0e24efe5a3083ef7e37d556c0ef","target":"graph","created_at":"2026-07-05T09:44:41Z","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/2412.03822/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) are increasingly deployed via public-facing interfaces to interact with millions of users, each with diverse preferences. Despite this, preference tuning of LLMs predominantly relies on reward models trained using binary judgments where annotators select the preferred choice out of pairs of model outputs. In this work, we argue that this reliance on binary choices does not capture the broader, aggregate preferences of the target user in real-world tasks. We propose a taxonomy that identifies two dimensions of subjectivity where different users disagree on the prefe","authors_text":"Chuanyang Jin, Hannah Rose Kirk, He He, Vishakh Padmakumar","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-05T02:35:46Z","title":"Beyond the Binary: Capturing Diverse Preferences With Reward Regularization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.03822","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":"sha256:7c98fc5f03194eb0e3289f53147d42ff07efbb7ec1b425da175d77488699984a","target":"record","created_at":"2026-07-05T09:44:41Z","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":"73d2f87e9ec1c17fa4a766c648162869ef068b0f5076e82f932ffeecf2464ad7","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-05T02:35:46Z","title_canon_sha256":"6fef43bb798a0b30b4fee70c7b608c80a5136d052a4314acde8d2552edcb9ae8"},"schema_version":"1.0","source":{"id":"2412.03822","kind":"arxiv","version":1}},"canonical_sha256":"2162b5ceaa3fa36a8d33acf58ce7716c71edd82cde0a6a4022e79c99c2a89020","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2162b5ceaa3fa36a8d33acf58ce7716c71edd82cde0a6a4022e79c99c2a89020","first_computed_at":"2026-07-05T09:44:41.779539Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:44:41.779539Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"H9MausHrtD7dyvFrFh71eAw+vujF8c08xoKBdTtDGNVOguzrC1WOEaoqGbSmIME4Wpwt7530w0VFtdoJd6zrAw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:44:41.780005Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.03822","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7c98fc5f03194eb0e3289f53147d42ff07efbb7ec1b425da175d77488699984a","sha256:4fe41596bbd2c465c231b2badd4d10bceb0bd0e24efe5a3083ef7e37d556c0ef"],"state_sha256":"c3881a928b469c77f787b313bd6bb94d4d464cd8fa1878788917f490584524ee"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ASqyHbbQbTEacj182lFN1G3DJFZFDr/5bvmnOpVtKMMDP3XtKoLJb/5lRl51MDE5BN84YuU84tAz9+/d7EVlCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-12T22:46:19.838179Z","bundle_sha256":"839c50a063c5342d14b513433ad77889a2c20cec3d2a61f4f99b445b203d00e3"}}