{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:OKH46CSTMQZHVZRQXBDWSU7TBR","short_pith_number":"pith:OKH46CST","canonical_record":{"source":{"id":"2402.11436","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-18T03:10:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"0988abd496c98f1ca530571187985f3c5aeddeb9f19454250fbe334f621e1994","abstract_canon_sha256":"3120c849ccfe0cedb0cf8dbfc5ab0dee9bfbb03f583f33799833c9166e313729"},"schema_version":"1.0"},"canonical_sha256":"728fcf0a5364327ae630b8476953f30c575bf9a322f74a8670a4984902c481bb","source":{"kind":"arxiv","id":"2402.11436","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.11436","created_at":"2026-07-05T08:33:35Z"},{"alias_kind":"arxiv_version","alias_value":"2402.11436v2","created_at":"2026-07-05T08:33:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.11436","created_at":"2026-07-05T08:33:35Z"},{"alias_kind":"pith_short_12","alias_value":"OKH46CSTMQZH","created_at":"2026-07-05T08:33:35Z"},{"alias_kind":"pith_short_16","alias_value":"OKH46CSTMQZHVZRQ","created_at":"2026-07-05T08:33:35Z"},{"alias_kind":"pith_short_8","alias_value":"OKH46CST","created_at":"2026-07-05T08:33:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:OKH46CSTMQZHVZRQXBDWSU7TBR","target":"record","payload":{"canonical_record":{"source":{"id":"2402.11436","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-18T03:10:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"0988abd496c98f1ca530571187985f3c5aeddeb9f19454250fbe334f621e1994","abstract_canon_sha256":"3120c849ccfe0cedb0cf8dbfc5ab0dee9bfbb03f583f33799833c9166e313729"},"schema_version":"1.0"},"canonical_sha256":"728fcf0a5364327ae630b8476953f30c575bf9a322f74a8670a4984902c481bb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:33:35.802594Z","signature_b64":"lj368ZSGQMdjfbBWP9ynR3LTwgg72QkPYWgOvWiQL67cfAvyaQ0lhU/Eg97E7t1Z6okZf2MjTEcYiE4amlcmBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"728fcf0a5364327ae630b8476953f30c575bf9a322f74a8670a4984902c481bb","last_reissued_at":"2026-07-05T08:33:35.802006Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:33:35.802006Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2402.11436","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":"2026-07-05T08:33:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5HZKaBsoCFsr8NqTIT3f94NYFZd1LIDXcbH2Sm/VyEII8uAFs7IKdItALegXBEBDb9KfeX6ZOsyCB+k/bArMDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T15:57:04.487089Z"},"content_sha256":"d4d5fdf9fcb75f823f0c55d2333db98e2af224f1c226698a2f48d1f39bbea230","schema_version":"1.0","event_id":"sha256:d4d5fdf9fcb75f823f0c55d2333db98e2af224f1c226698a2f48d1f39bbea230"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:OKH46CSTMQZHVZRQXBDWSU7TBR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Pride and Prejudice: LLM Amplifies Self-Bias in Self-Refinement","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Guanglei Zhu, Lei Li, Liangming Pan, Wenda Xu, William Yang Wang, Xuandong Zhao","submitted_at":"2024-02-18T03:10:39Z","abstract_excerpt":"Recent studies show that large language models (LLMs) improve their performance through self-feedback on certain tasks while degrade on others. We discovered that such a contrary is due to LLM's bias in evaluating their own output. In this paper, we formally define LLM's self-bias - the tendency to favor its own generation - using two statistics. We analyze six LLMs (GPT-4, GPT-3.5, Gemini, LLaMA2, Mixtral and DeepSeek) on translation, constrained text generation, and mathematical reasoning tasks. We find that self-bias is prevalent in all examined LLMs across multiple languages and tasks. Our"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.11436","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/2402.11436/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":"2026-07-05T08:33:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FlXDTZbNLKyfE3vUl0X6DNRPOpNzabafHcNFr32tgncoaLDp1UZfH96ASm65mdqxgsVzsHM8+ttHZUdX0StCBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-15T15:57:04.487617Z"},"content_sha256":"b28ce67425aad037210a2f7424b5a0b1cf25d8b2fb117306cdc46f9d0967793a","schema_version":"1.0","event_id":"sha256:b28ce67425aad037210a2f7424b5a0b1cf25d8b2fb117306cdc46f9d0967793a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/OKH46CSTMQZHVZRQXBDWSU7TBR/bundle.json","state_url":"https://pith.science/pith/OKH46CSTMQZHVZRQXBDWSU7TBR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/OKH46CSTMQZHVZRQXBDWSU7TBR/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":"2026-08-15T15:57:04Z","links":{"resolver":"https://pith.science/pith/OKH46CSTMQZHVZRQXBDWSU7TBR","bundle":"https://pith.science/pith/OKH46CSTMQZHVZRQXBDWSU7TBR/bundle.json","state":"https://pith.science/pith/OKH46CSTMQZHVZRQXBDWSU7TBR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/OKH46CSTMQZHVZRQXBDWSU7TBR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:OKH46CSTMQZHVZRQXBDWSU7TBR","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":"3120c849ccfe0cedb0cf8dbfc5ab0dee9bfbb03f583f33799833c9166e313729","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-18T03:10:39Z","title_canon_sha256":"0988abd496c98f1ca530571187985f3c5aeddeb9f19454250fbe334f621e1994"},"schema_version":"1.0","source":{"id":"2402.11436","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.11436","created_at":"2026-07-05T08:33:35Z"},{"alias_kind":"arxiv_version","alias_value":"2402.11436v2","created_at":"2026-07-05T08:33:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.11436","created_at":"2026-07-05T08:33:35Z"},{"alias_kind":"pith_short_12","alias_value":"OKH46CSTMQZH","created_at":"2026-07-05T08:33:35Z"},{"alias_kind":"pith_short_16","alias_value":"OKH46CSTMQZHVZRQ","created_at":"2026-07-05T08:33:35Z"},{"alias_kind":"pith_short_8","alias_value":"OKH46CST","created_at":"2026-07-05T08:33:35Z"}],"graph_snapshots":[{"event_id":"sha256:b28ce67425aad037210a2f7424b5a0b1cf25d8b2fb117306cdc46f9d0967793a","target":"graph","created_at":"2026-07-05T08:33:35Z","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/2402.11436/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent studies show that large language models (LLMs) improve their performance through self-feedback on certain tasks while degrade on others. We discovered that such a contrary is due to LLM's bias in evaluating their own output. In this paper, we formally define LLM's self-bias - the tendency to favor its own generation - using two statistics. We analyze six LLMs (GPT-4, GPT-3.5, Gemini, LLaMA2, Mixtral and DeepSeek) on translation, constrained text generation, and mathematical reasoning tasks. We find that self-bias is prevalent in all examined LLMs across multiple languages and tasks. Our","authors_text":"Guanglei Zhu, Lei Li, Liangming Pan, Wenda Xu, William Yang Wang, Xuandong Zhao","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-18T03:10:39Z","title":"Pride and Prejudice: LLM Amplifies Self-Bias in Self-Refinement"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.11436","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:d4d5fdf9fcb75f823f0c55d2333db98e2af224f1c226698a2f48d1f39bbea230","target":"record","created_at":"2026-07-05T08:33:35Z","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":"3120c849ccfe0cedb0cf8dbfc5ab0dee9bfbb03f583f33799833c9166e313729","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-18T03:10:39Z","title_canon_sha256":"0988abd496c98f1ca530571187985f3c5aeddeb9f19454250fbe334f621e1994"},"schema_version":"1.0","source":{"id":"2402.11436","kind":"arxiv","version":2}},"canonical_sha256":"728fcf0a5364327ae630b8476953f30c575bf9a322f74a8670a4984902c481bb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"728fcf0a5364327ae630b8476953f30c575bf9a322f74a8670a4984902c481bb","first_computed_at":"2026-07-05T08:33:35.802006Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:33:35.802006Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"lj368ZSGQMdjfbBWP9ynR3LTwgg72QkPYWgOvWiQL67cfAvyaQ0lhU/Eg97E7t1Z6okZf2MjTEcYiE4amlcmBg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:33:35.802594Z","signed_message":"canonical_sha256_bytes"},"source_id":"2402.11436","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d4d5fdf9fcb75f823f0c55d2333db98e2af224f1c226698a2f48d1f39bbea230","sha256:b28ce67425aad037210a2f7424b5a0b1cf25d8b2fb117306cdc46f9d0967793a"],"state_sha256":"d83185e3314cebb2f72b9ae553e6687fbb24830959820002746ff3e2dc6c5689"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"D4/o5YjHlaMNEFr6RTUNos9QVu9udwYnwRBsofZoKxAPYYj6sKomcCVIhUrehc87Be+ujiII4Kh4SPl1t49PDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-15T15:57:04.492132Z","bundle_sha256":"745eebb5d41eac4e87284c7f806a2064c9b6bd7cef9e6db68a5e8e882d94bce1"}}