{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:USSZWKAXIJL5EDI5VSOKVOS4OZ","short_pith_number":"pith:USSZWKAX","canonical_record":{"source":{"id":"2501.18243","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.AP","submitted_at":"2025-01-30T10:21:10Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"467bc99d653789ce6d204b2201f0ea20a033132bba892b5a8720d586027aaacc","abstract_canon_sha256":"671e43e4f2ca902a34b9559c35ed19f90b646892a9b6be93cc92c0da0dcff7e8"},"schema_version":"1.0"},"canonical_sha256":"a4a59b28174257d20d1dac9caaba5c7664998befcced0df81b3b614b3701fea8","source":{"kind":"arxiv","id":"2501.18243","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.18243","created_at":"2026-07-05T10:07:31Z"},{"alias_kind":"arxiv_version","alias_value":"2501.18243v1","created_at":"2026-07-05T10:07:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.18243","created_at":"2026-07-05T10:07:31Z"},{"alias_kind":"pith_short_12","alias_value":"USSZWKAXIJL5","created_at":"2026-07-05T10:07:31Z"},{"alias_kind":"pith_short_16","alias_value":"USSZWKAXIJL5EDI5","created_at":"2026-07-05T10:07:31Z"},{"alias_kind":"pith_short_8","alias_value":"USSZWKAX","created_at":"2026-07-05T10:07:31Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:USSZWKAXIJL5EDI5VSOKVOS4OZ","target":"record","payload":{"canonical_record":{"source":{"id":"2501.18243","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.AP","submitted_at":"2025-01-30T10:21:10Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"467bc99d653789ce6d204b2201f0ea20a033132bba892b5a8720d586027aaacc","abstract_canon_sha256":"671e43e4f2ca902a34b9559c35ed19f90b646892a9b6be93cc92c0da0dcff7e8"},"schema_version":"1.0"},"canonical_sha256":"a4a59b28174257d20d1dac9caaba5c7664998befcced0df81b3b614b3701fea8","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:07:31.579364Z","signature_b64":"Byq0+yHQeEItSyRqifs7apHq3mdJmieDQXM4iAVweRnJOL3Q/Ts2BbbbjM5faWJkydv/qo2oDRB6bnIyjRpODg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a4a59b28174257d20d1dac9caaba5c7664998befcced0df81b3b614b3701fea8","last_reissued_at":"2026-07-05T10:07:31.578858Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:07:31.578858Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2501.18243","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":"2026-07-05T10:07:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ottBxpq31AcJyWDQlFBncId7b0DR9MVPHmdPDcfOPEB5lmd82Hzsm94s2Xt/ck4/BqJan+JDAOgarFgF4OwBCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T21:54:48.261314Z"},"content_sha256":"4de71bd68c5530a77562fa6e3b9ace38c2bfb93fa11510e9d9e27f993b54b885","schema_version":"1.0","event_id":"sha256:4de71bd68c5530a77562fa6e3b9ace38c2bfb93fa11510e9d9e27f993b54b885"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:USSZWKAXIJL5EDI5VSOKVOS4OZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Statistical multi-metric evaluation and visualization of LLM system predictive performance","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"stat.AP","authors_text":"Assaf Toledo, Eitan Farchi, Orna Raz, Samuel Ackerman","submitted_at":"2025-01-30T10:21:10Z","abstract_excerpt":"The evaluation of generative or discriminative large language model (LLM)-based systems is often a complex multi-dimensional problem. Typically, a set of system configuration alternatives are evaluated on one or more benchmark datasets, each with one or more evaluation metrics, which may differ between datasets. We often want to evaluate -- with a statistical measure of significance -- whether systems perform differently either on a given dataset according to a single metric, on aggregate across metrics on a dataset, or across datasets. Such evaluations can be done to support decision-making, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.18243","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/2501.18243/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":"2026-07-05T10:07:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gx9A+bt7651TZnYL+L3MFFbY5LQgS5rZUOwTW99W2og7NHZHZoPpUZm2XkSZLa+ljEHHpzBMhM3Zj5vLmao2BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T21:54:48.261836Z"},"content_sha256":"4310accddea0ae7cfd61f15831c739074b8c6eac5839f6180c8547fa4b516dda","schema_version":"1.0","event_id":"sha256:4310accddea0ae7cfd61f15831c739074b8c6eac5839f6180c8547fa4b516dda"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/USSZWKAXIJL5EDI5VSOKVOS4OZ/bundle.json","state_url":"https://pith.science/pith/USSZWKAXIJL5EDI5VSOKVOS4OZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/USSZWKAXIJL5EDI5VSOKVOS4OZ/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":"2026-08-10T21:54:48Z","links":{"resolver":"https://pith.science/pith/USSZWKAXIJL5EDI5VSOKVOS4OZ","bundle":"https://pith.science/pith/USSZWKAXIJL5EDI5VSOKVOS4OZ/bundle.json","state":"https://pith.science/pith/USSZWKAXIJL5EDI5VSOKVOS4OZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/USSZWKAXIJL5EDI5VSOKVOS4OZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:USSZWKAXIJL5EDI5VSOKVOS4OZ","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":"671e43e4f2ca902a34b9559c35ed19f90b646892a9b6be93cc92c0da0dcff7e8","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.AP","submitted_at":"2025-01-30T10:21:10Z","title_canon_sha256":"467bc99d653789ce6d204b2201f0ea20a033132bba892b5a8720d586027aaacc"},"schema_version":"1.0","source":{"id":"2501.18243","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.18243","created_at":"2026-07-05T10:07:31Z"},{"alias_kind":"arxiv_version","alias_value":"2501.18243v1","created_at":"2026-07-05T10:07:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.18243","created_at":"2026-07-05T10:07:31Z"},{"alias_kind":"pith_short_12","alias_value":"USSZWKAXIJL5","created_at":"2026-07-05T10:07:31Z"},{"alias_kind":"pith_short_16","alias_value":"USSZWKAXIJL5EDI5","created_at":"2026-07-05T10:07:31Z"},{"alias_kind":"pith_short_8","alias_value":"USSZWKAX","created_at":"2026-07-05T10:07:31Z"}],"graph_snapshots":[{"event_id":"sha256:4310accddea0ae7cfd61f15831c739074b8c6eac5839f6180c8547fa4b516dda","target":"graph","created_at":"2026-07-05T10:07:31Z","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/2501.18243/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The evaluation of generative or discriminative large language model (LLM)-based systems is often a complex multi-dimensional problem. Typically, a set of system configuration alternatives are evaluated on one or more benchmark datasets, each with one or more evaluation metrics, which may differ between datasets. We often want to evaluate -- with a statistical measure of significance -- whether systems perform differently either on a given dataset according to a single metric, on aggregate across metrics on a dataset, or across datasets. Such evaluations can be done to support decision-making, ","authors_text":"Assaf Toledo, Eitan Farchi, Orna Raz, Samuel Ackerman","cross_cats":["cs.CL","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.AP","submitted_at":"2025-01-30T10:21:10Z","title":"Statistical multi-metric evaluation and visualization of LLM system predictive performance"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.18243","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":"sha256:4de71bd68c5530a77562fa6e3b9ace38c2bfb93fa11510e9d9e27f993b54b885","target":"record","created_at":"2026-07-05T10:07:31Z","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":"671e43e4f2ca902a34b9559c35ed19f90b646892a9b6be93cc92c0da0dcff7e8","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.AP","submitted_at":"2025-01-30T10:21:10Z","title_canon_sha256":"467bc99d653789ce6d204b2201f0ea20a033132bba892b5a8720d586027aaacc"},"schema_version":"1.0","source":{"id":"2501.18243","kind":"arxiv","version":1}},"canonical_sha256":"a4a59b28174257d20d1dac9caaba5c7664998befcced0df81b3b614b3701fea8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a4a59b28174257d20d1dac9caaba5c7664998befcced0df81b3b614b3701fea8","first_computed_at":"2026-07-05T10:07:31.578858Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:07:31.578858Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Byq0+yHQeEItSyRqifs7apHq3mdJmieDQXM4iAVweRnJOL3Q/Ts2BbbbjM5faWJkydv/qo2oDRB6bnIyjRpODg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:07:31.579364Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.18243","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4de71bd68c5530a77562fa6e3b9ace38c2bfb93fa11510e9d9e27f993b54b885","sha256:4310accddea0ae7cfd61f15831c739074b8c6eac5839f6180c8547fa4b516dda"],"state_sha256":"51baf0c2cdecdc966d0717c941b6fa7076ad0387b48a8276ab5261597ed58871"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7NFEEwA2sUykXSqLc9ZG1CAterInB6NUzIZBpiPoLt/KkdeCBWTuGgoxs1CVX0Jfu/aLoGYXIfaWQL9aHoOpCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T21:54:48.267672Z","bundle_sha256":"63ffe6e28dd4c6185f1cb7fb14ac7bffcfffe4da3795834b220adcd3ae1c7f85"}}