{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:4EUSQFUICRDAJTE523NUCBOK5P","short_pith_number":"pith:4EUSQFUI","canonical_record":{"source":{"id":"2508.11847","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2025-08-16T00:01:33Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"724ae8d3ba35d6001c83b11ae2136f32b9d08b3fab5b62c5f2533d0515a1cc11","abstract_canon_sha256":"0f776ad5cb327a6c324c7ed98595e17d2aad26f2357b79f00b59665460a33384"},"schema_version":"1.0"},"canonical_sha256":"e129281688144604cc9dd6db4105caebd580d9ebd340cd7a0de18bdef1975717","source":{"kind":"arxiv","id":"2508.11847","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.11847","created_at":"2026-07-10T00:18:41Z"},{"alias_kind":"arxiv_version","alias_value":"2508.11847v4","created_at":"2026-07-10T00:18:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.11847","created_at":"2026-07-10T00:18:41Z"},{"alias_kind":"pith_short_12","alias_value":"4EUSQFUICRDA","created_at":"2026-07-10T00:18:41Z"},{"alias_kind":"pith_short_16","alias_value":"4EUSQFUICRDAJTE5","created_at":"2026-07-10T00:18:41Z"},{"alias_kind":"pith_short_8","alias_value":"4EUSQFUI","created_at":"2026-07-10T00:18:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:4EUSQFUICRDAJTE523NUCBOK5P","target":"record","payload":{"canonical_record":{"source":{"id":"2508.11847","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2025-08-16T00:01:33Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"724ae8d3ba35d6001c83b11ae2136f32b9d08b3fab5b62c5f2533d0515a1cc11","abstract_canon_sha256":"0f776ad5cb327a6c324c7ed98595e17d2aad26f2357b79f00b59665460a33384"},"schema_version":"1.0"},"canonical_sha256":"e129281688144604cc9dd6db4105caebd580d9ebd340cd7a0de18bdef1975717","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-10T00:18:41.059538Z","signature_b64":"qSrxL484K4c/REfW+kQq796GxECzy2UUz3XAG2eWdBEyPl4OtQ5f124oytEL8svyJMoI3pkRpwwGFgv9vQ6ZDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e129281688144604cc9dd6db4105caebd580d9ebd340cd7a0de18bdef1975717","last_reissued_at":"2026-07-10T00:18:41.059043Z","signature_status":"signed_v1","first_computed_at":"2026-07-10T00:18:41.059043Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2508.11847","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-10T00:18:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fWMvKG7uP7CMuInkanYBrbPCcoTAsBufACO4V8P4AnPLW0KNUW7fvpZ7HlTcZYef7OK3D/HLudPr2mg58QihCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T01:01:13.606029Z"},"content_sha256":"e3791844237ea3923b58401a63d6ef2c811e019ffa210820b8ee35753859c6bd","schema_version":"1.0","event_id":"sha256:e3791844237ea3923b58401a63d6ef2c811e019ffa210820b8ee35753859c6bd"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:4EUSQFUICRDAJTE523NUCBOK5P","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Dropping Just a Handful of Preferences Can Change Top Large Language Model Rankings","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Dennis Wei, Jenny Y. Huang, Tamara Broderick, Yunyi Shen","submitted_at":"2025-08-16T00:01:33Z","abstract_excerpt":"We propose a method for evaluating the robustness of widely used LLM ranking systems -- variants of a Bradley--Terry model -- to dropping a worst-case very small fraction of preference data. Our approach is computationally fast and easy to adopt. When we apply our method to matchups from popular LLM ranking platforms, including Chatbot Arena and derivatives, we find that the rankings of top-performing models can be remarkably sensitive to the removal of a small fraction of preferences; for instance, dropping just 0.003% of human preferences can change the top-ranked model on Chatbot Arena. Our"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.11847","kind":"arxiv","version":4},"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/2508.11847/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":"2026-07-10T00:18:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0P9FsuuZ2GBSTn3OW29WrZuNedIiszq/8Uhz12ip2t483hRJC+f3whaP2a9gn7nqDNHeSbfJwBe5YQem6bMkAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T01:01:13.606575Z"},"content_sha256":"9ed86c56aa57f3b10fe2804c00a1f13fe3ac04d4a1f53b1e1dda9373a0cefb68","schema_version":"1.0","event_id":"sha256:9ed86c56aa57f3b10fe2804c00a1f13fe3ac04d4a1f53b1e1dda9373a0cefb68"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/4EUSQFUICRDAJTE523NUCBOK5P/bundle.json","state_url":"https://pith.science/pith/4EUSQFUICRDAJTE523NUCBOK5P/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/4EUSQFUICRDAJTE523NUCBOK5P/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":"2026-08-16T01:01:13Z","links":{"resolver":"https://pith.science/pith/4EUSQFUICRDAJTE523NUCBOK5P","bundle":"https://pith.science/pith/4EUSQFUICRDAJTE523NUCBOK5P/bundle.json","state":"https://pith.science/pith/4EUSQFUICRDAJTE523NUCBOK5P/state.json","well_known_bundle":"https://pith.science/.well-known/pith/4EUSQFUICRDAJTE523NUCBOK5P/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:4EUSQFUICRDAJTE523NUCBOK5P","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":"0f776ad5cb327a6c324c7ed98595e17d2aad26f2357b79f00b59665460a33384","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2025-08-16T00:01:33Z","title_canon_sha256":"724ae8d3ba35d6001c83b11ae2136f32b9d08b3fab5b62c5f2533d0515a1cc11"},"schema_version":"1.0","source":{"id":"2508.11847","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.11847","created_at":"2026-07-10T00:18:41Z"},{"alias_kind":"arxiv_version","alias_value":"2508.11847v4","created_at":"2026-07-10T00:18:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.11847","created_at":"2026-07-10T00:18:41Z"},{"alias_kind":"pith_short_12","alias_value":"4EUSQFUICRDA","created_at":"2026-07-10T00:18:41Z"},{"alias_kind":"pith_short_16","alias_value":"4EUSQFUICRDAJTE5","created_at":"2026-07-10T00:18:41Z"},{"alias_kind":"pith_short_8","alias_value":"4EUSQFUI","created_at":"2026-07-10T00:18:41Z"}],"graph_snapshots":[{"event_id":"sha256:9ed86c56aa57f3b10fe2804c00a1f13fe3ac04d4a1f53b1e1dda9373a0cefb68","target":"graph","created_at":"2026-07-10T00:18:41Z","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/2508.11847/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We propose a method for evaluating the robustness of widely used LLM ranking systems -- variants of a Bradley--Terry model -- to dropping a worst-case very small fraction of preference data. Our approach is computationally fast and easy to adopt. When we apply our method to matchups from popular LLM ranking platforms, including Chatbot Arena and derivatives, we find that the rankings of top-performing models can be remarkably sensitive to the removal of a small fraction of preferences; for instance, dropping just 0.003% of human preferences can change the top-ranked model on Chatbot Arena. Our","authors_text":"Dennis Wei, Jenny Y. Huang, Tamara Broderick, Yunyi Shen","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2025-08-16T00:01:33Z","title":"Dropping Just a Handful of Preferences Can Change Top Large Language Model Rankings"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.11847","kind":"arxiv","version":4},"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:e3791844237ea3923b58401a63d6ef2c811e019ffa210820b8ee35753859c6bd","target":"record","created_at":"2026-07-10T00:18:41Z","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":"0f776ad5cb327a6c324c7ed98595e17d2aad26f2357b79f00b59665460a33384","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2025-08-16T00:01:33Z","title_canon_sha256":"724ae8d3ba35d6001c83b11ae2136f32b9d08b3fab5b62c5f2533d0515a1cc11"},"schema_version":"1.0","source":{"id":"2508.11847","kind":"arxiv","version":4}},"canonical_sha256":"e129281688144604cc9dd6db4105caebd580d9ebd340cd7a0de18bdef1975717","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e129281688144604cc9dd6db4105caebd580d9ebd340cd7a0de18bdef1975717","first_computed_at":"2026-07-10T00:18:41.059043Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-10T00:18:41.059043Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"qSrxL484K4c/REfW+kQq796GxECzy2UUz3XAG2eWdBEyPl4OtQ5f124oytEL8svyJMoI3pkRpwwGFgv9vQ6ZDg==","signature_status":"signed_v1","signed_at":"2026-07-10T00:18:41.059538Z","signed_message":"canonical_sha256_bytes"},"source_id":"2508.11847","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e3791844237ea3923b58401a63d6ef2c811e019ffa210820b8ee35753859c6bd","sha256:9ed86c56aa57f3b10fe2804c00a1f13fe3ac04d4a1f53b1e1dda9373a0cefb68"],"state_sha256":"24b18ae53d9b6e9cb7949709a512b888d299e3ce7f0c984003d512fa369cb5c0"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VFkcM3oxeKn9Ezx96QOQRw1+8WzTOsiX4/ZcZDbIAOAf3gAVSyyNOtz+0X+jhoK4tJ0UD0VDBz/QVbt0q9IHBQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-16T01:01:13.611490Z","bundle_sha256":"0c560ed2b9935529533fc3b7026ff3169c413847624e4f1d1193b52f36995fd8"}}