{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:TKEHXPH37EG2XXCB2YMRUEOWBQ","short_pith_number":"pith:TKEHXPH3","canonical_record":{"source":{"id":"2501.07493","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-13T17:12:38Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"e5f6ffa8ea2f8d37235d4d0898c436cf5289f8a804a63a36f3f0ca3fe4286a84","abstract_canon_sha256":"da4a4cf7007641a1d1a28f09f75762f9dae22f58efc3b204e49164a190e9362f"},"schema_version":"1.0"},"canonical_sha256":"9a887bbcfbf90dabdc41d6191a11d60c1b793d2ab1c732d3855540305b4c683a","source":{"kind":"arxiv","id":"2501.07493","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.07493","created_at":"2026-07-05T10:00:29Z"},{"alias_kind":"arxiv_version","alias_value":"2501.07493v1","created_at":"2026-07-05T10:00:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.07493","created_at":"2026-07-05T10:00:29Z"},{"alias_kind":"pith_short_12","alias_value":"TKEHXPH37EG2","created_at":"2026-07-05T10:00:29Z"},{"alias_kind":"pith_short_16","alias_value":"TKEHXPH37EG2XXCB","created_at":"2026-07-05T10:00:29Z"},{"alias_kind":"pith_short_8","alias_value":"TKEHXPH3","created_at":"2026-07-05T10:00:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:TKEHXPH37EG2XXCB2YMRUEOWBQ","target":"record","payload":{"canonical_record":{"source":{"id":"2501.07493","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-13T17:12:38Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"e5f6ffa8ea2f8d37235d4d0898c436cf5289f8a804a63a36f3f0ca3fe4286a84","abstract_canon_sha256":"da4a4cf7007641a1d1a28f09f75762f9dae22f58efc3b204e49164a190e9362f"},"schema_version":"1.0"},"canonical_sha256":"9a887bbcfbf90dabdc41d6191a11d60c1b793d2ab1c732d3855540305b4c683a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:00:29.311203Z","signature_b64":"x32KLFw1fjqqZ6bQM0VDatl/NQkdu6QjIxpjqEZ1Zm9PZoHIyjeAwSvLOLtNjATRt6yHXv4tlYm7goPAFBB6Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9a887bbcfbf90dabdc41d6191a11d60c1b793d2ab1c732d3855540305b4c683a","last_reissued_at":"2026-07-05T10:00:29.310703Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:00:29.310703Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2501.07493","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":"2026-07-05T10:00:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"svP6Dd9B0p4r/gYg6c8NrpkGkntpJ7s7T8P4gIF+cxYv3in4MOcA9Zl+fGGP3bGBJ/q9ZordnSYMeKeVxFLwCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T04:36:04.101574Z"},"content_sha256":"da46548f6072bcf23dd346a02c548a01e8f01a53b92b9f7b02482999a5804178","schema_version":"1.0","event_id":"sha256:da46548f6072bcf23dd346a02c548a01e8f01a53b92b9f7b02482999a5804178"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:TKEHXPH37EG2XXCB2YMRUEOWBQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Exploring and Mitigating Adversarial Manipulation of Voting-Based Leaderboards","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Anastasios Angelopoulos, Chiyuan Zhang, Christopher A. Choquette-Choo, Daphne Ippolito, Florian Tramer, Ion Stoica, Katherine Lee, Ken Ziyu Liu, Matthew Jagielski, Milad Nasr, Nicholas Carlini, Wei-Lin Chiang, Yangsibo Huang","submitted_at":"2025-01-13T17:12:38Z","abstract_excerpt":"It is now common to evaluate Large Language Models (LLMs) by having humans manually vote to evaluate model outputs, in contrast to typical benchmarks that evaluate knowledge or skill at some particular task. Chatbot Arena, the most popular benchmark of this type, ranks models by asking users to select the better response between two randomly selected models (without revealing which model was responsible for the generations). These platforms are widely trusted as a fair and accurate measure of LLM capabilities. In this paper, we show that if bot protection and other defenses are not implemented"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.07493","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/2501.07493/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":"2026-07-05T10:00:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NZuQw+8xumxvVZ0XtqzgqDwtV6EM9OS2fEjsasJpAGjO6XkS7U19d7StDqTPFn8kfIWlg+x6IFNb8wyDzla5Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-11T04:36:04.102098Z"},"content_sha256":"5705237bd0a466f44c4e370f1b212137e62154b42f9a501d9f92fdfc0bf4b785","schema_version":"1.0","event_id":"sha256:5705237bd0a466f44c4e370f1b212137e62154b42f9a501d9f92fdfc0bf4b785"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/TKEHXPH37EG2XXCB2YMRUEOWBQ/bundle.json","state_url":"https://pith.science/pith/TKEHXPH37EG2XXCB2YMRUEOWBQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/TKEHXPH37EG2XXCB2YMRUEOWBQ/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":"2026-08-11T04:36:04Z","links":{"resolver":"https://pith.science/pith/TKEHXPH37EG2XXCB2YMRUEOWBQ","bundle":"https://pith.science/pith/TKEHXPH37EG2XXCB2YMRUEOWBQ/bundle.json","state":"https://pith.science/pith/TKEHXPH37EG2XXCB2YMRUEOWBQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/TKEHXPH37EG2XXCB2YMRUEOWBQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:TKEHXPH37EG2XXCB2YMRUEOWBQ","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":"da4a4cf7007641a1d1a28f09f75762f9dae22f58efc3b204e49164a190e9362f","cross_cats_sorted":["cs.CR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-13T17:12:38Z","title_canon_sha256":"e5f6ffa8ea2f8d37235d4d0898c436cf5289f8a804a63a36f3f0ca3fe4286a84"},"schema_version":"1.0","source":{"id":"2501.07493","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.07493","created_at":"2026-07-05T10:00:29Z"},{"alias_kind":"arxiv_version","alias_value":"2501.07493v1","created_at":"2026-07-05T10:00:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.07493","created_at":"2026-07-05T10:00:29Z"},{"alias_kind":"pith_short_12","alias_value":"TKEHXPH37EG2","created_at":"2026-07-05T10:00:29Z"},{"alias_kind":"pith_short_16","alias_value":"TKEHXPH37EG2XXCB","created_at":"2026-07-05T10:00:29Z"},{"alias_kind":"pith_short_8","alias_value":"TKEHXPH3","created_at":"2026-07-05T10:00:29Z"}],"graph_snapshots":[{"event_id":"sha256:5705237bd0a466f44c4e370f1b212137e62154b42f9a501d9f92fdfc0bf4b785","target":"graph","created_at":"2026-07-05T10:00:29Z","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/2501.07493/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"It is now common to evaluate Large Language Models (LLMs) by having humans manually vote to evaluate model outputs, in contrast to typical benchmarks that evaluate knowledge or skill at some particular task. Chatbot Arena, the most popular benchmark of this type, ranks models by asking users to select the better response between two randomly selected models (without revealing which model was responsible for the generations). These platforms are widely trusted as a fair and accurate measure of LLM capabilities. In this paper, we show that if bot protection and other defenses are not implemented","authors_text":"Anastasios Angelopoulos, Chiyuan Zhang, Christopher A. Choquette-Choo, Daphne Ippolito, Florian Tramer, Ion Stoica, Katherine Lee, Ken Ziyu Liu, Matthew Jagielski, Milad Nasr, Nicholas Carlini, Wei-Lin Chiang, Yangsibo Huang","cross_cats":["cs.CR"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-13T17:12:38Z","title":"Exploring and Mitigating Adversarial Manipulation of Voting-Based Leaderboards"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.07493","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:da46548f6072bcf23dd346a02c548a01e8f01a53b92b9f7b02482999a5804178","target":"record","created_at":"2026-07-05T10:00:29Z","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":"da4a4cf7007641a1d1a28f09f75762f9dae22f58efc3b204e49164a190e9362f","cross_cats_sorted":["cs.CR"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-13T17:12:38Z","title_canon_sha256":"e5f6ffa8ea2f8d37235d4d0898c436cf5289f8a804a63a36f3f0ca3fe4286a84"},"schema_version":"1.0","source":{"id":"2501.07493","kind":"arxiv","version":1}},"canonical_sha256":"9a887bbcfbf90dabdc41d6191a11d60c1b793d2ab1c732d3855540305b4c683a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9a887bbcfbf90dabdc41d6191a11d60c1b793d2ab1c732d3855540305b4c683a","first_computed_at":"2026-07-05T10:00:29.310703Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:00:29.310703Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"x32KLFw1fjqqZ6bQM0VDatl/NQkdu6QjIxpjqEZ1Zm9PZoHIyjeAwSvLOLtNjATRt6yHXv4tlYm7goPAFBB6Dg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:00:29.311203Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.07493","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:da46548f6072bcf23dd346a02c548a01e8f01a53b92b9f7b02482999a5804178","sha256:5705237bd0a466f44c4e370f1b212137e62154b42f9a501d9f92fdfc0bf4b785"],"state_sha256":"b41c0ab6caf160643d8ad6d4a168c9035f7b46e9e5436ad8131df256411e07d6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Z6FAgSBz/PqdRvenGlZw3p5nB4GuWTSi3EuMuB1Nwb8gVLBTHeiU1OYPzApUTt5gYb6Wg8A4juVttYsIa286BA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-11T04:36:04.105519Z","bundle_sha256":"a52d10b1656e14bc3b4989db14b2c2db2d7c2dae77d917216b41f11d58956d1a"}}