{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:ZLUTRSCPWNGQ5R7VX3SBZPTH6N","short_pith_number":"pith:ZLUTRSCP","canonical_record":{"source":{"id":"2605.30916","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T07:01:38Z","cross_cats_sorted":["cs.GT","econ.TH"],"title_canon_sha256":"99ca4d8388a64094050c418da85b3cb494a1fedfb38eee17d2d1eaf21322f439","abstract_canon_sha256":"0fb088cea39875565c708a10d74a3669a7aa8593f48087a894ea415677a29353"},"schema_version":"1.0"},"canonical_sha256":"cae938c84fb34d0ec7f5bee41cbe67f352ef38b44b55b0b5866c6fe31752d4e2","source":{"kind":"arxiv","id":"2605.30916","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.30916","created_at":"2026-06-01T01:03:25Z"},{"alias_kind":"arxiv_version","alias_value":"2605.30916v1","created_at":"2026-06-01T01:03:25Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30916","created_at":"2026-06-01T01:03:25Z"},{"alias_kind":"pith_short_12","alias_value":"ZLUTRSCPWNGQ","created_at":"2026-06-01T01:03:25Z"},{"alias_kind":"pith_short_16","alias_value":"ZLUTRSCPWNGQ5R7V","created_at":"2026-06-01T01:03:25Z"},{"alias_kind":"pith_short_8","alias_value":"ZLUTRSCP","created_at":"2026-06-01T01:03:25Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:ZLUTRSCPWNGQ5R7VX3SBZPTH6N","target":"record","payload":{"canonical_record":{"source":{"id":"2605.30916","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T07:01:38Z","cross_cats_sorted":["cs.GT","econ.TH"],"title_canon_sha256":"99ca4d8388a64094050c418da85b3cb494a1fedfb38eee17d2d1eaf21322f439","abstract_canon_sha256":"0fb088cea39875565c708a10d74a3669a7aa8593f48087a894ea415677a29353"},"schema_version":"1.0"},"canonical_sha256":"cae938c84fb34d0ec7f5bee41cbe67f352ef38b44b55b0b5866c6fe31752d4e2","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-01T01:03:25.704168Z","signature_b64":"IMyBzgnUQ5i6FZ3b3O/smEjY2MO1ztkROSOYPTkrAu6jiKIr8G+g2IENx5zyvfTjjFtQIXCOlsQgXznTNU35AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cae938c84fb34d0ec7f5bee41cbe67f352ef38b44b55b0b5866c6fe31752d4e2","last_reissued_at":"2026-06-01T01:03:25.703369Z","signature_status":"signed_v1","first_computed_at":"2026-06-01T01:03:25.703369Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.30916","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":"2026-06-01T01:03:25Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5lK99oREDL1Z9oe74TjJ3XBZ4eJt5RFqrC8bYDgHZ/Ldm5/CUxunBh25YQTfLs+q/dulkdgYKIYIYgyR45K9AA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-30T09:34:21.285817Z"},"content_sha256":"198326ce033b4ccb8861b7ca4db4785278bbf645b78c9cae0c2295db520527c3","schema_version":"1.0","event_id":"sha256:198326ce033b4ccb8861b7ca4db4785278bbf645b78c9cae0c2295db520527c3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:ZLUTRSCPWNGQ5R7VX3SBZPTH6N","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Welfare, Improvability, and Variance: A Principal-Agent Approach to Optimal Benchmark Item Aggregation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.GT","econ.TH"],"primary_cat":"cs.LG","authors_text":"Andreas Haupt, Anka Reuel, Justin Hartenstein, Mykel Kochenderfer, Sanmi Koyejo","submitted_at":"2026-05-29T07:01:38Z","abstract_excerpt":"AI benchmarks have well-documented limitations, with prior work examining contamination, saturation, and construct underspecification. Aggregation has received far less attention: benchmarks are typically summarized by uniformly averaging item-level scores, implicitly treating every test item as equally valuable. We model benchmarking as a multitask principal-agent game and show that the welfare loss from a benchmark is determined jointly by three item-level primitives: alignment with normative welfare priorities, marginal improvability, and performance variance. We translate the theory into a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30916","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/2605.30916/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":"2026-06-01T01:03:25Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hV96fkW4Btg2yefJHzV10H0Hop92HsqPTIYHGung5qlG8q7gG5RktdYGWEN9pxWZZtXi9snvrfab7AKWIGpiCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-30T09:34:21.286202Z"},"content_sha256":"8e5e4600b4c995745608b6077e1ddd744bba5a6c3e48cd26a27478f8c4d9e739","schema_version":"1.0","event_id":"sha256:8e5e4600b4c995745608b6077e1ddd744bba5a6c3e48cd26a27478f8c4d9e739"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZLUTRSCPWNGQ5R7VX3SBZPTH6N/bundle.json","state_url":"https://pith.science/pith/ZLUTRSCPWNGQ5R7VX3SBZPTH6N/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZLUTRSCPWNGQ5R7VX3SBZPTH6N/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":"2026-06-30T09:34:21Z","links":{"resolver":"https://pith.science/pith/ZLUTRSCPWNGQ5R7VX3SBZPTH6N","bundle":"https://pith.science/pith/ZLUTRSCPWNGQ5R7VX3SBZPTH6N/bundle.json","state":"https://pith.science/pith/ZLUTRSCPWNGQ5R7VX3SBZPTH6N/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZLUTRSCPWNGQ5R7VX3SBZPTH6N/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:ZLUTRSCPWNGQ5R7VX3SBZPTH6N","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":"0fb088cea39875565c708a10d74a3669a7aa8593f48087a894ea415677a29353","cross_cats_sorted":["cs.GT","econ.TH"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T07:01:38Z","title_canon_sha256":"99ca4d8388a64094050c418da85b3cb494a1fedfb38eee17d2d1eaf21322f439"},"schema_version":"1.0","source":{"id":"2605.30916","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.30916","created_at":"2026-06-01T01:03:25Z"},{"alias_kind":"arxiv_version","alias_value":"2605.30916v1","created_at":"2026-06-01T01:03:25Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30916","created_at":"2026-06-01T01:03:25Z"},{"alias_kind":"pith_short_12","alias_value":"ZLUTRSCPWNGQ","created_at":"2026-06-01T01:03:25Z"},{"alias_kind":"pith_short_16","alias_value":"ZLUTRSCPWNGQ5R7V","created_at":"2026-06-01T01:03:25Z"},{"alias_kind":"pith_short_8","alias_value":"ZLUTRSCP","created_at":"2026-06-01T01:03:25Z"}],"graph_snapshots":[{"event_id":"sha256:8e5e4600b4c995745608b6077e1ddd744bba5a6c3e48cd26a27478f8c4d9e739","target":"graph","created_at":"2026-06-01T01:03:25Z","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/2605.30916/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"AI benchmarks have well-documented limitations, with prior work examining contamination, saturation, and construct underspecification. Aggregation has received far less attention: benchmarks are typically summarized by uniformly averaging item-level scores, implicitly treating every test item as equally valuable. We model benchmarking as a multitask principal-agent game and show that the welfare loss from a benchmark is determined jointly by three item-level primitives: alignment with normative welfare priorities, marginal improvability, and performance variance. We translate the theory into a","authors_text":"Andreas Haupt, Anka Reuel, Justin Hartenstein, Mykel Kochenderfer, Sanmi Koyejo","cross_cats":["cs.GT","econ.TH"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T07:01:38Z","title":"Welfare, Improvability, and Variance: A Principal-Agent Approach to Optimal Benchmark Item Aggregation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30916","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":"sha256:198326ce033b4ccb8861b7ca4db4785278bbf645b78c9cae0c2295db520527c3","target":"record","created_at":"2026-06-01T01:03:25Z","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":"0fb088cea39875565c708a10d74a3669a7aa8593f48087a894ea415677a29353","cross_cats_sorted":["cs.GT","econ.TH"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T07:01:38Z","title_canon_sha256":"99ca4d8388a64094050c418da85b3cb494a1fedfb38eee17d2d1eaf21322f439"},"schema_version":"1.0","source":{"id":"2605.30916","kind":"arxiv","version":1}},"canonical_sha256":"cae938c84fb34d0ec7f5bee41cbe67f352ef38b44b55b0b5866c6fe31752d4e2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"cae938c84fb34d0ec7f5bee41cbe67f352ef38b44b55b0b5866c6fe31752d4e2","first_computed_at":"2026-06-01T01:03:25.703369Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-01T01:03:25.703369Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"IMyBzgnUQ5i6FZ3b3O/smEjY2MO1ztkROSOYPTkrAu6jiKIr8G+g2IENx5zyvfTjjFtQIXCOlsQgXznTNU35AQ==","signature_status":"signed_v1","signed_at":"2026-06-01T01:03:25.704168Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.30916","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:198326ce033b4ccb8861b7ca4db4785278bbf645b78c9cae0c2295db520527c3","sha256:8e5e4600b4c995745608b6077e1ddd744bba5a6c3e48cd26a27478f8c4d9e739"],"state_sha256":"d0abc3a453652f077d6aed187cc32991ea1e85f25ab7fd229cbb67cfbf8f3052"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oINMSn5zTn2SZFF50bBXLe8Kw84gpoutL2R/xY8sdZeNvmUhZkCaD2SvIlEeo8c7f0Fs8CRmexXMxvQ/TZ+CCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-30T09:34:21.288264Z","bundle_sha256":"4f2cd8ba32dd10c6648c34adc9c36821add6f6068769db899b8817efa109accc"}}