{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:3TTIUMGLYKHUZHT2HUHLDC7GSR","short_pith_number":"pith:3TTIUMGL","canonical_record":{"source":{"id":"2506.17111","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-06-20T16:11:25Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"18e13cdf3c9d3db2591af7db80e7eca1bf1df75f07a5bc1406f89c7049c161d2","abstract_canon_sha256":"cf25cb3181f152f70f65528273f923e8d712a7ce79e220983c9bc5bc0ab9803f"},"schema_version":"1.0"},"canonical_sha256":"dce68a30cbc28f4c9e7a3d0eb18be6944ae96b2fdb2f6413de7aa1182d91cd88","source":{"kind":"arxiv","id":"2506.17111","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.17111","created_at":"2026-07-05T11:24:56Z"},{"alias_kind":"arxiv_version","alias_value":"2506.17111v1","created_at":"2026-07-05T11:24:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.17111","created_at":"2026-07-05T11:24:56Z"},{"alias_kind":"pith_short_12","alias_value":"3TTIUMGLYKHU","created_at":"2026-07-05T11:24:56Z"},{"alias_kind":"pith_short_16","alias_value":"3TTIUMGLYKHUZHT2","created_at":"2026-07-05T11:24:56Z"},{"alias_kind":"pith_short_8","alias_value":"3TTIUMGL","created_at":"2026-07-05T11:24:56Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:3TTIUMGLYKHUZHT2HUHLDC7GSR","target":"record","payload":{"canonical_record":{"source":{"id":"2506.17111","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-06-20T16:11:25Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"18e13cdf3c9d3db2591af7db80e7eca1bf1df75f07a5bc1406f89c7049c161d2","abstract_canon_sha256":"cf25cb3181f152f70f65528273f923e8d712a7ce79e220983c9bc5bc0ab9803f"},"schema_version":"1.0"},"canonical_sha256":"dce68a30cbc28f4c9e7a3d0eb18be6944ae96b2fdb2f6413de7aa1182d91cd88","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:24:56.473729Z","signature_b64":"UulM5QWRNInRIrzwxi9D9cmDfnxTBqdSRST/Hv9BCfCZSHzaDLR4rE5t4FmOhDBVci31TF86fbG+ltkgcJePBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dce68a30cbc28f4c9e7a3d0eb18be6944ae96b2fdb2f6413de7aa1182d91cd88","last_reissued_at":"2026-07-05T11:24:56.473313Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:24:56.473313Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2506.17111","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":"2026-07-05T11:24:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"X4VI4+ozkp97vg9xiKHObRcOznOqnOgYfNBHxRu5AbQhuxwgRDruchu99xRg2uNioKUxqfXcvjtTNFqX2g8YCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T09:14:39.215365Z"},"content_sha256":"c43f56c0a901eb1cc19d94a2d2f5d5e2ce7332f5d39e9b1bbc82bcdfc1b714db","schema_version":"1.0","event_id":"sha256:c43f56c0a901eb1cc19d94a2d2f5d5e2ce7332f5d39e9b1bbc82bcdfc1b714db"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:3TTIUMGLYKHUZHT2HUHLDC7GSR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Are Bias Evaluation Methods Biased ?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Ioana Giurgiu, Lina Berrayana, Luis Garc\\'es-Erice, Sean Rooney","submitted_at":"2025-06-20T16:11:25Z","abstract_excerpt":"The creation of benchmarks to evaluate the safety of Large Language Models is one of the key activities within the trusted AI community. These benchmarks allow models to be compared for different aspects of safety such as toxicity, bias, harmful behavior etc. Independent benchmarks adopt different approaches with distinct data sets and evaluation methods. We investigate how robust such benchmarks are by using different approaches to rank a set of representative models for bias and compare how similar are the overall rankings. We show that different but widely used bias evaluations methods resu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.17111","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/2506.17111/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":"2026-07-05T11:24:56Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tCv7NepUKGSyjjVoZfdqau1nKw2UZx/xH/xCYoJcD0WjfvqfKyLdTFk8tlRxyds1jW9efmCZOVAPdYynrPDDDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T09:14:39.219951Z"},"content_sha256":"60b818589777671a42e222795905a5d6b3a8212e4e2925bddd56f0e788f766a7","schema_version":"1.0","event_id":"sha256:60b818589777671a42e222795905a5d6b3a8212e4e2925bddd56f0e788f766a7"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3TTIUMGLYKHUZHT2HUHLDC7GSR/bundle.json","state_url":"https://pith.science/pith/3TTIUMGLYKHUZHT2HUHLDC7GSR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3TTIUMGLYKHUZHT2HUHLDC7GSR/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":"2026-08-04T09:14:39Z","links":{"resolver":"https://pith.science/pith/3TTIUMGLYKHUZHT2HUHLDC7GSR","bundle":"https://pith.science/pith/3TTIUMGLYKHUZHT2HUHLDC7GSR/bundle.json","state":"https://pith.science/pith/3TTIUMGLYKHUZHT2HUHLDC7GSR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3TTIUMGLYKHUZHT2HUHLDC7GSR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:3TTIUMGLYKHUZHT2HUHLDC7GSR","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":"cf25cb3181f152f70f65528273f923e8d712a7ce79e220983c9bc5bc0ab9803f","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-06-20T16:11:25Z","title_canon_sha256":"18e13cdf3c9d3db2591af7db80e7eca1bf1df75f07a5bc1406f89c7049c161d2"},"schema_version":"1.0","source":{"id":"2506.17111","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.17111","created_at":"2026-07-05T11:24:56Z"},{"alias_kind":"arxiv_version","alias_value":"2506.17111v1","created_at":"2026-07-05T11:24:56Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.17111","created_at":"2026-07-05T11:24:56Z"},{"alias_kind":"pith_short_12","alias_value":"3TTIUMGLYKHU","created_at":"2026-07-05T11:24:56Z"},{"alias_kind":"pith_short_16","alias_value":"3TTIUMGLYKHUZHT2","created_at":"2026-07-05T11:24:56Z"},{"alias_kind":"pith_short_8","alias_value":"3TTIUMGL","created_at":"2026-07-05T11:24:56Z"}],"graph_snapshots":[{"event_id":"sha256:60b818589777671a42e222795905a5d6b3a8212e4e2925bddd56f0e788f766a7","target":"graph","created_at":"2026-07-05T11:24:56Z","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/2506.17111/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The creation of benchmarks to evaluate the safety of Large Language Models is one of the key activities within the trusted AI community. These benchmarks allow models to be compared for different aspects of safety such as toxicity, bias, harmful behavior etc. Independent benchmarks adopt different approaches with distinct data sets and evaluation methods. We investigate how robust such benchmarks are by using different approaches to rank a set of representative models for bias and compare how similar are the overall rankings. We show that different but widely used bias evaluations methods resu","authors_text":"Ioana Giurgiu, Lina Berrayana, Luis Garc\\'es-Erice, Sean Rooney","cross_cats":["cs.CL"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-06-20T16:11:25Z","title":"Are Bias Evaluation Methods Biased ?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.17111","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:c43f56c0a901eb1cc19d94a2d2f5d5e2ce7332f5d39e9b1bbc82bcdfc1b714db","target":"record","created_at":"2026-07-05T11:24:56Z","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":"cf25cb3181f152f70f65528273f923e8d712a7ce79e220983c9bc5bc0ab9803f","cross_cats_sorted":["cs.CL"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-06-20T16:11:25Z","title_canon_sha256":"18e13cdf3c9d3db2591af7db80e7eca1bf1df75f07a5bc1406f89c7049c161d2"},"schema_version":"1.0","source":{"id":"2506.17111","kind":"arxiv","version":1}},"canonical_sha256":"dce68a30cbc28f4c9e7a3d0eb18be6944ae96b2fdb2f6413de7aa1182d91cd88","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dce68a30cbc28f4c9e7a3d0eb18be6944ae96b2fdb2f6413de7aa1182d91cd88","first_computed_at":"2026-07-05T11:24:56.473313Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:24:56.473313Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"UulM5QWRNInRIrzwxi9D9cmDfnxTBqdSRST/Hv9BCfCZSHzaDLR4rE5t4FmOhDBVci31TF86fbG+ltkgcJePBw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:24:56.473729Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.17111","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c43f56c0a901eb1cc19d94a2d2f5d5e2ce7332f5d39e9b1bbc82bcdfc1b714db","sha256:60b818589777671a42e222795905a5d6b3a8212e4e2925bddd56f0e788f766a7"],"state_sha256":"842d8f0299eb23fb0ca29e9d139406de7656203b241e15eb3e3aad56b3137f7e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ew1Fse96q7T9HUM2B96O2lw1QfmrLfR9w3XSqbyCrErMO7LB/XLRPnMvrOni96zNTQJf77QCc0iV2lUXvQQ4CA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T09:14:39.227485Z","bundle_sha256":"0e2664d05e8c1b31a62e44aab3df9fa86fd475d253688484758560df820e50e7"}}