{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:QK6J3RXDT2I4UY2HQMWXB7TPTR","short_pith_number":"pith:QK6J3RXD","canonical_record":{"source":{"id":"2407.01085","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-07-01T08:37:41Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"3435626d3901f2e8c75813816c8452ca0eb8bb3845a3e5e9d89988ccbd5079e7","abstract_canon_sha256":"447d05a74a29fd18cd7da4541bac9e9bb34eb83d089bc77d70b4a172dbc2b459"},"schema_version":"1.0"},"canonical_sha256":"82bc9dc6e39e91ca6347832d70fe6f9c4077cc06fb9aa94e9b7e51f4fc39a738","source":{"kind":"arxiv","id":"2407.01085","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.01085","created_at":"2026-07-05T12:04:27Z"},{"alias_kind":"arxiv_version","alias_value":"2407.01085v5","created_at":"2026-07-05T12:04:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.01085","created_at":"2026-07-05T12:04:27Z"},{"alias_kind":"pith_short_12","alias_value":"QK6J3RXDT2I4","created_at":"2026-07-05T12:04:27Z"},{"alias_kind":"pith_short_16","alias_value":"QK6J3RXDT2I4UY2H","created_at":"2026-07-05T12:04:27Z"},{"alias_kind":"pith_short_8","alias_value":"QK6J3RXD","created_at":"2026-07-05T12:04:27Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:QK6J3RXDT2I4UY2HQMWXB7TPTR","target":"record","payload":{"canonical_record":{"source":{"id":"2407.01085","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-07-01T08:37:41Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"3435626d3901f2e8c75813816c8452ca0eb8bb3845a3e5e9d89988ccbd5079e7","abstract_canon_sha256":"447d05a74a29fd18cd7da4541bac9e9bb34eb83d089bc77d70b4a172dbc2b459"},"schema_version":"1.0"},"canonical_sha256":"82bc9dc6e39e91ca6347832d70fe6f9c4077cc06fb9aa94e9b7e51f4fc39a738","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:04:27.284255Z","signature_b64":"wIxQ6QDZmbAVfenYFENDYzYcImCvgN96mzqr5cK2j4d9lvDmtcSLFQEkdz5stQfyAhT0I5fVUDStuYoCYdYnCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"82bc9dc6e39e91ca6347832d70fe6f9c4077cc06fb9aa94e9b7e51f4fc39a738","last_reissued_at":"2026-07-05T12:04:27.283745Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:04:27.283745Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2407.01085","source_version":5,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T12:04:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"UtXvw/ykXNZignPSPjCbKZfxWHKw5FUHZxJF9e4b8Ebu1iUWt8xmBBhkASZ9XoQxCTNebAnrWp5jSMVEMasQDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-21T09:12:25.628602Z"},"content_sha256":"e0cc111ab98cb537bdd5f8dd6deed66458bdd65ca60ebf4d536f74e833f72591","schema_version":"1.0","event_id":"sha256:e0cc111ab98cb537bdd5f8dd6deed66458bdd65ca60ebf4d536f74e833f72591"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:QK6J3RXDT2I4UY2HQMWXB7TPTR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Explaining Length Bias in LLM-Based Preference Evaluations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Hui Xiong, Jianxun Lian, Jieyu Zhang, Kaize Ding, Linxin Song, Nicholas Jing Yuan, Tianfu Wang, Zhengyu Chen, Zhengyu Hu, Zheyuan Xiao","submitted_at":"2024-07-01T08:37:41Z","abstract_excerpt":"The use of large language models (LLMs) as judges, particularly in preference comparisons, has become widespread, but this reveals a notable bias towards longer responses, undermining the reliability of such evaluations. To better understand such bias, we propose to decompose the preference evaluation metric, specifically the win rate, into two key components: desirability and information mass, where the former is length-independent and related to trustworthiness such as correctness, toxicity, and consistency, and the latter is length-dependent and represents the amount of information in the r"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.01085","kind":"arxiv","version":5},"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/2407.01085/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":"2026-07-05T12:04:27Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gHZLXWO8aYDJgZQAiy7n+iEQQmWUwpbFqirlqglpfDsg6XZtiNDRN4hKChd4vb53AD4H0wwiO6iyQW31ff8MCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-21T09:12:25.629319Z"},"content_sha256":"f6ba0c10a55e811f01a04121223aebe714424b2dbad1b5bea4715d26461a66c1","schema_version":"1.0","event_id":"sha256:f6ba0c10a55e811f01a04121223aebe714424b2dbad1b5bea4715d26461a66c1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/QK6J3RXDT2I4UY2HQMWXB7TPTR/bundle.json","state_url":"https://pith.science/pith/QK6J3RXDT2I4UY2HQMWXB7TPTR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/QK6J3RXDT2I4UY2HQMWXB7TPTR/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":"2026-08-21T09:12:25Z","links":{"resolver":"https://pith.science/pith/QK6J3RXDT2I4UY2HQMWXB7TPTR","bundle":"https://pith.science/pith/QK6J3RXDT2I4UY2HQMWXB7TPTR/bundle.json","state":"https://pith.science/pith/QK6J3RXDT2I4UY2HQMWXB7TPTR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/QK6J3RXDT2I4UY2HQMWXB7TPTR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:QK6J3RXDT2I4UY2HQMWXB7TPTR","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":"447d05a74a29fd18cd7da4541bac9e9bb34eb83d089bc77d70b4a172dbc2b459","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-07-01T08:37:41Z","title_canon_sha256":"3435626d3901f2e8c75813816c8452ca0eb8bb3845a3e5e9d89988ccbd5079e7"},"schema_version":"1.0","source":{"id":"2407.01085","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2407.01085","created_at":"2026-07-05T12:04:27Z"},{"alias_kind":"arxiv_version","alias_value":"2407.01085v5","created_at":"2026-07-05T12:04:27Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.01085","created_at":"2026-07-05T12:04:27Z"},{"alias_kind":"pith_short_12","alias_value":"QK6J3RXDT2I4","created_at":"2026-07-05T12:04:27Z"},{"alias_kind":"pith_short_16","alias_value":"QK6J3RXDT2I4UY2H","created_at":"2026-07-05T12:04:27Z"},{"alias_kind":"pith_short_8","alias_value":"QK6J3RXD","created_at":"2026-07-05T12:04:27Z"}],"graph_snapshots":[{"event_id":"sha256:f6ba0c10a55e811f01a04121223aebe714424b2dbad1b5bea4715d26461a66c1","target":"graph","created_at":"2026-07-05T12:04:27Z","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/2407.01085/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The use of large language models (LLMs) as judges, particularly in preference comparisons, has become widespread, but this reveals a notable bias towards longer responses, undermining the reliability of such evaluations. To better understand such bias, we propose to decompose the preference evaluation metric, specifically the win rate, into two key components: desirability and information mass, where the former is length-independent and related to trustworthiness such as correctness, toxicity, and consistency, and the latter is length-dependent and represents the amount of information in the r","authors_text":"Hui Xiong, Jianxun Lian, Jieyu Zhang, Kaize Ding, Linxin Song, Nicholas Jing Yuan, Tianfu Wang, Zhengyu Chen, Zhengyu Hu, Zheyuan Xiao","cross_cats":["cs.CL"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-07-01T08:37:41Z","title":"Explaining Length Bias in LLM-Based Preference Evaluations"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.01085","kind":"arxiv","version":5},"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:e0cc111ab98cb537bdd5f8dd6deed66458bdd65ca60ebf4d536f74e833f72591","target":"record","created_at":"2026-07-05T12:04:27Z","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":"447d05a74a29fd18cd7da4541bac9e9bb34eb83d089bc77d70b4a172dbc2b459","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-07-01T08:37:41Z","title_canon_sha256":"3435626d3901f2e8c75813816c8452ca0eb8bb3845a3e5e9d89988ccbd5079e7"},"schema_version":"1.0","source":{"id":"2407.01085","kind":"arxiv","version":5}},"canonical_sha256":"82bc9dc6e39e91ca6347832d70fe6f9c4077cc06fb9aa94e9b7e51f4fc39a738","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"82bc9dc6e39e91ca6347832d70fe6f9c4077cc06fb9aa94e9b7e51f4fc39a738","first_computed_at":"2026-07-05T12:04:27.283745Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T12:04:27.283745Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"wIxQ6QDZmbAVfenYFENDYzYcImCvgN96mzqr5cK2j4d9lvDmtcSLFQEkdz5stQfyAhT0I5fVUDStuYoCYdYnCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T12:04:27.284255Z","signed_message":"canonical_sha256_bytes"},"source_id":"2407.01085","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e0cc111ab98cb537bdd5f8dd6deed66458bdd65ca60ebf4d536f74e833f72591","sha256:f6ba0c10a55e811f01a04121223aebe714424b2dbad1b5bea4715d26461a66c1"],"state_sha256":"4e86ef8bcb806647daf5709d51c6db3cfd52e46f7457fce63057a212fb7753ad"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"sy+yYw3ZhEmMJeuBZeW0wBiHG5stEhRl6KAVlfd9vnGsW3EoFYYH+iaeeWGVZkIDVdSylt/+vuU2NQHw0yRjDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-21T09:12:25.641159Z","bundle_sha256":"077673a80dff9ec7d47407c59c7ae7f5f1ceb88797e5f4783a4a51708f367637"}}