{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:E2AL5ZW6KE4TSTVYYYM7KE3I2S","short_pith_number":"pith:E2AL5ZW6","canonical_record":{"source":{"id":"2606.02547","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.GT","submitted_at":"2026-06-01T17:49:02Z","cross_cats_sorted":[],"title_canon_sha256":"119266e4bd42b65a78e689988406b36477cb97e7b89de0fca3463aaa1918c105","abstract_canon_sha256":"37619be02481407aa3a5a2b3a95fb55206a164314dc80e20e0cdd0448c20591a"},"schema_version":"1.0"},"canonical_sha256":"2680bee6de5139394eb8c619f51368d4a8130720dbb89246f5fa8f6e3a77bd34","source":{"kind":"arxiv","id":"2606.02547","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.02547","created_at":"2026-06-02T03:05:09Z"},{"alias_kind":"arxiv_version","alias_value":"2606.02547v1","created_at":"2026-06-02T03:05:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.02547","created_at":"2026-06-02T03:05:09Z"},{"alias_kind":"pith_short_12","alias_value":"E2AL5ZW6KE4T","created_at":"2026-06-02T03:05:09Z"},{"alias_kind":"pith_short_16","alias_value":"E2AL5ZW6KE4TSTVY","created_at":"2026-06-02T03:05:09Z"},{"alias_kind":"pith_short_8","alias_value":"E2AL5ZW6","created_at":"2026-06-02T03:05:09Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:E2AL5ZW6KE4TSTVYYYM7KE3I2S","target":"record","payload":{"canonical_record":{"source":{"id":"2606.02547","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.GT","submitted_at":"2026-06-01T17:49:02Z","cross_cats_sorted":[],"title_canon_sha256":"119266e4bd42b65a78e689988406b36477cb97e7b89de0fca3463aaa1918c105","abstract_canon_sha256":"37619be02481407aa3a5a2b3a95fb55206a164314dc80e20e0cdd0448c20591a"},"schema_version":"1.0"},"canonical_sha256":"2680bee6de5139394eb8c619f51368d4a8130720dbb89246f5fa8f6e3a77bd34","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T03:05:09.302426Z","signature_b64":"0hNBPNXCzvq9jQEuDOLmi5WaawnWyC8XOZe1qsmfHpita/nB8lQxPRUawvVZzDg6Hw5oGELCOgE69PnCOP8SCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2680bee6de5139394eb8c619f51368d4a8130720dbb89246f5fa8f6e3a77bd34","last_reissued_at":"2026-06-02T03:05:09.302017Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T03:05:09.302017Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.02547","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":"2026-06-02T03:05:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"U05tIbVbCVLAXyiQ4a4ptKE0R1Wgjczjyk9ghaFs5vn0xqIHNfiF/v/GCShKv4+yvO3QzQj5jlCP0yS8NFK1DA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-03T16:39:13.814706Z"},"content_sha256":"d2926592115bcd4597e3126a6fae9692f9f17ada9ce96feef2faf13b9d53a1ff","schema_version":"1.0","event_id":"sha256:d2926592115bcd4597e3126a6fae9692f9f17ada9ce96feef2faf13b9d53a1ff"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:E2AL5ZW6KE4TSTVYYYM7KE3I2S","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Pluralistic Leaderboards","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.GT","authors_text":"Ariel D. Procaccia, Han Shao, Kunhe Yang, Nika Haghtalab, Serena Lutong Wang","submitted_at":"2026-06-01T17:49:02Z","abstract_excerpt":"Recent leaderboard-based evaluations of large language models aggregate user feedback by fitting a Bradley--Terry model to pairwise comparisons, producing a single global ranking based on a latent quality score. While appealing for its simplicity, this approach is incompatible with heterogeneous preferences: when LLMs are used across diverse tasks and use cases, users who favor fundamentally different model behaviors can be systematically misrepresented when collapsed into a single quality score. To address this issue, we study \\emph{pluralistic leaderboards} that aim to remain \\emph{stable} w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.02547","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/2606.02547/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":"2026-06-02T03:05:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jSlu/KQ+9hQOXv4v4jVQ9G0cZKk4jclAwQu3kDYHIa6JZWFQ61ykmL2zLQpxmqGOU9RSqsohVWH12jxzvwAUAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-03T16:39:13.815071Z"},"content_sha256":"02a9f6201aaaf1a441f67f3fe6fedd5f9a0a6392de17275abe813d935c3411bc","schema_version":"1.0","event_id":"sha256:02a9f6201aaaf1a441f67f3fe6fedd5f9a0a6392de17275abe813d935c3411bc"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/E2AL5ZW6KE4TSTVYYYM7KE3I2S/bundle.json","state_url":"https://pith.science/pith/E2AL5ZW6KE4TSTVYYYM7KE3I2S/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/E2AL5ZW6KE4TSTVYYYM7KE3I2S/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":"2026-06-03T16:39:13Z","links":{"resolver":"https://pith.science/pith/E2AL5ZW6KE4TSTVYYYM7KE3I2S","bundle":"https://pith.science/pith/E2AL5ZW6KE4TSTVYYYM7KE3I2S/bundle.json","state":"https://pith.science/pith/E2AL5ZW6KE4TSTVYYYM7KE3I2S/state.json","well_known_bundle":"https://pith.science/.well-known/pith/E2AL5ZW6KE4TSTVYYYM7KE3I2S/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:E2AL5ZW6KE4TSTVYYYM7KE3I2S","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":"37619be02481407aa3a5a2b3a95fb55206a164314dc80e20e0cdd0448c20591a","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.GT","submitted_at":"2026-06-01T17:49:02Z","title_canon_sha256":"119266e4bd42b65a78e689988406b36477cb97e7b89de0fca3463aaa1918c105"},"schema_version":"1.0","source":{"id":"2606.02547","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.02547","created_at":"2026-06-02T03:05:09Z"},{"alias_kind":"arxiv_version","alias_value":"2606.02547v1","created_at":"2026-06-02T03:05:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.02547","created_at":"2026-06-02T03:05:09Z"},{"alias_kind":"pith_short_12","alias_value":"E2AL5ZW6KE4T","created_at":"2026-06-02T03:05:09Z"},{"alias_kind":"pith_short_16","alias_value":"E2AL5ZW6KE4TSTVY","created_at":"2026-06-02T03:05:09Z"},{"alias_kind":"pith_short_8","alias_value":"E2AL5ZW6","created_at":"2026-06-02T03:05:09Z"}],"graph_snapshots":[{"event_id":"sha256:02a9f6201aaaf1a441f67f3fe6fedd5f9a0a6392de17275abe813d935c3411bc","target":"graph","created_at":"2026-06-02T03:05:09Z","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/2606.02547/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent leaderboard-based evaluations of large language models aggregate user feedback by fitting a Bradley--Terry model to pairwise comparisons, producing a single global ranking based on a latent quality score. While appealing for its simplicity, this approach is incompatible with heterogeneous preferences: when LLMs are used across diverse tasks and use cases, users who favor fundamentally different model behaviors can be systematically misrepresented when collapsed into a single quality score. To address this issue, we study \\emph{pluralistic leaderboards} that aim to remain \\emph{stable} w","authors_text":"Ariel D. Procaccia, Han Shao, Kunhe Yang, Nika Haghtalab, Serena Lutong Wang","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.GT","submitted_at":"2026-06-01T17:49:02Z","title":"Pluralistic Leaderboards"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.02547","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:d2926592115bcd4597e3126a6fae9692f9f17ada9ce96feef2faf13b9d53a1ff","target":"record","created_at":"2026-06-02T03:05:09Z","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":"37619be02481407aa3a5a2b3a95fb55206a164314dc80e20e0cdd0448c20591a","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.GT","submitted_at":"2026-06-01T17:49:02Z","title_canon_sha256":"119266e4bd42b65a78e689988406b36477cb97e7b89de0fca3463aaa1918c105"},"schema_version":"1.0","source":{"id":"2606.02547","kind":"arxiv","version":1}},"canonical_sha256":"2680bee6de5139394eb8c619f51368d4a8130720dbb89246f5fa8f6e3a77bd34","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2680bee6de5139394eb8c619f51368d4a8130720dbb89246f5fa8f6e3a77bd34","first_computed_at":"2026-06-02T03:05:09.302017Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-02T03:05:09.302017Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"0hNBPNXCzvq9jQEuDOLmi5WaawnWyC8XOZe1qsmfHpita/nB8lQxPRUawvVZzDg6Hw5oGELCOgE69PnCOP8SCg==","signature_status":"signed_v1","signed_at":"2026-06-02T03:05:09.302426Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.02547","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d2926592115bcd4597e3126a6fae9692f9f17ada9ce96feef2faf13b9d53a1ff","sha256:02a9f6201aaaf1a441f67f3fe6fedd5f9a0a6392de17275abe813d935c3411bc"],"state_sha256":"d64044cdcda6e0f2979c7fd620b65be931128e382140007a1fcc851650b2eea1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hbXR++irbKd3/clgNE4wuSidcEPf+azpdKDSOwTGM/f1rb2dxu5cqW1L8Y0TPEElWB3e2iBIOJc/E1+OzNdbAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-03T16:39:13.817063Z","bundle_sha256":"6f318a070623dca84f196f47b668f19adb755c3d4b5d98b8892426822b427022"}}