{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:I4BEE7V6NG6T5GOXUZFOGG5LSL","short_pith_number":"pith:I4BEE7V6","canonical_record":{"source":{"id":"2508.15835","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-08-19T03:59:04Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a280e408e8e0177ac250d3c64aece3fa9a6bc78aa75ae4425203d47348259d19","abstract_canon_sha256":"8520c06f02d51088ca65f0496a66d24d693fd5372ccb521164b21ad0cc78e803"},"schema_version":"1.0"},"canonical_sha256":"4702427ebe69bd3e99d7a64ae31bab92d1210b41c9d3a44ff842bf5bceb13c8b","source":{"kind":"arxiv","id":"2508.15835","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.15835","created_at":"2026-07-05T11:57:21Z"},{"alias_kind":"arxiv_version","alias_value":"2508.15835v1","created_at":"2026-07-05T11:57:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.15835","created_at":"2026-07-05T11:57:21Z"},{"alias_kind":"pith_short_12","alias_value":"I4BEE7V6NG6T","created_at":"2026-07-05T11:57:21Z"},{"alias_kind":"pith_short_16","alias_value":"I4BEE7V6NG6T5GOX","created_at":"2026-07-05T11:57:21Z"},{"alias_kind":"pith_short_8","alias_value":"I4BEE7V6","created_at":"2026-07-05T11:57:21Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:I4BEE7V6NG6T5GOXUZFOGG5LSL","target":"record","payload":{"canonical_record":{"source":{"id":"2508.15835","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-08-19T03:59:04Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a280e408e8e0177ac250d3c64aece3fa9a6bc78aa75ae4425203d47348259d19","abstract_canon_sha256":"8520c06f02d51088ca65f0496a66d24d693fd5372ccb521164b21ad0cc78e803"},"schema_version":"1.0"},"canonical_sha256":"4702427ebe69bd3e99d7a64ae31bab92d1210b41c9d3a44ff842bf5bceb13c8b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:57:21.732692Z","signature_b64":"jbTdQzno5KdQWEcdn59MiMvTBT04Xbl4sg0jxCmGBQOh/Xx+rSAzhy9twwz8EVAhD29Wj6CbTDF0wLtsRocsDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4702427ebe69bd3e99d7a64ae31bab92d1210b41c9d3a44ff842bf5bceb13c8b","last_reissued_at":"2026-07-05T11:57:21.732254Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:57:21.732254Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2508.15835","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":"2026-07-05T11:57:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GYVx00emcinTX/JbD3IfWC7wLb1Ya9d04WCvFY7uUR06JupqMxEnoWzy2Mh0h0w89GwAv7hEYXR+/VxDcrnWBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T03:05:40.500592Z"},"content_sha256":"14675c3774052f58321e2800e064f3a7bf79f9b0f9bfaf46e829ba2cedf71409","schema_version":"1.0","event_id":"sha256:14675c3774052f58321e2800e064f3a7bf79f9b0f9bfaf46e829ba2cedf71409"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:I4BEE7V6NG6T5GOXUZFOGG5LSL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Alvorada-Bench: Can Language Models Solve Brazilian University Entrance Exams?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Henrique Godoy","submitted_at":"2025-08-19T03:59:04Z","abstract_excerpt":"Language models are increasingly used in Brazil, but most evaluation remains English-centric. This paper presents Alvorada-Bench, a 4,515-question, text-only benchmark drawn from five Brazilian university entrance examinations. Evaluating twenty models under zero-shot, role-playing, and chain-of-thought prompting, producing 270,900 responses with structured self-reports of confidence, perceived difficulty, and Bloom level. The top models exceed 94% accuracy overall, but accuracy declines on Mathematics and on the engineering oriented IME and ITA exams, indicating persistent weaknesses in multi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.15835","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/2508.15835/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":"2026-07-05T11:57:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6YnPmYXQp44udnNlfv5n9wzzy8hMT2UvcjJSAAk3CsWQvmBnZJMslWwCjadUYolwQBS/DfmLjZ6gEICVwU0NCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T03:05:40.501004Z"},"content_sha256":"6ebc98b10a1bd776f6cc60d753e5c8430a65cdbddaef7c6f0ba1fc3935207e8b","schema_version":"1.0","event_id":"sha256:6ebc98b10a1bd776f6cc60d753e5c8430a65cdbddaef7c6f0ba1fc3935207e8b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/I4BEE7V6NG6T5GOXUZFOGG5LSL/bundle.json","state_url":"https://pith.science/pith/I4BEE7V6NG6T5GOXUZFOGG5LSL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/I4BEE7V6NG6T5GOXUZFOGG5LSL/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":"2026-07-07T03:05:40Z","links":{"resolver":"https://pith.science/pith/I4BEE7V6NG6T5GOXUZFOGG5LSL","bundle":"https://pith.science/pith/I4BEE7V6NG6T5GOXUZFOGG5LSL/bundle.json","state":"https://pith.science/pith/I4BEE7V6NG6T5GOXUZFOGG5LSL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/I4BEE7V6NG6T5GOXUZFOGG5LSL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:I4BEE7V6NG6T5GOXUZFOGG5LSL","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":"8520c06f02d51088ca65f0496a66d24d693fd5372ccb521164b21ad0cc78e803","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-08-19T03:59:04Z","title_canon_sha256":"a280e408e8e0177ac250d3c64aece3fa9a6bc78aa75ae4425203d47348259d19"},"schema_version":"1.0","source":{"id":"2508.15835","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.15835","created_at":"2026-07-05T11:57:21Z"},{"alias_kind":"arxiv_version","alias_value":"2508.15835v1","created_at":"2026-07-05T11:57:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.15835","created_at":"2026-07-05T11:57:21Z"},{"alias_kind":"pith_short_12","alias_value":"I4BEE7V6NG6T","created_at":"2026-07-05T11:57:21Z"},{"alias_kind":"pith_short_16","alias_value":"I4BEE7V6NG6T5GOX","created_at":"2026-07-05T11:57:21Z"},{"alias_kind":"pith_short_8","alias_value":"I4BEE7V6","created_at":"2026-07-05T11:57:21Z"}],"graph_snapshots":[{"event_id":"sha256:6ebc98b10a1bd776f6cc60d753e5c8430a65cdbddaef7c6f0ba1fc3935207e8b","target":"graph","created_at":"2026-07-05T11:57:21Z","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/2508.15835/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Language models are increasingly used in Brazil, but most evaluation remains English-centric. This paper presents Alvorada-Bench, a 4,515-question, text-only benchmark drawn from five Brazilian university entrance examinations. Evaluating twenty models under zero-shot, role-playing, and chain-of-thought prompting, producing 270,900 responses with structured self-reports of confidence, perceived difficulty, and Bloom level. The top models exceed 94% accuracy overall, but accuracy declines on Mathematics and on the engineering oriented IME and ITA exams, indicating persistent weaknesses in multi","authors_text":"Henrique Godoy","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-08-19T03:59:04Z","title":"Alvorada-Bench: Can Language Models Solve Brazilian University Entrance Exams?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.15835","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":"sha256:14675c3774052f58321e2800e064f3a7bf79f9b0f9bfaf46e829ba2cedf71409","target":"record","created_at":"2026-07-05T11:57:21Z","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":"8520c06f02d51088ca65f0496a66d24d693fd5372ccb521164b21ad0cc78e803","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-08-19T03:59:04Z","title_canon_sha256":"a280e408e8e0177ac250d3c64aece3fa9a6bc78aa75ae4425203d47348259d19"},"schema_version":"1.0","source":{"id":"2508.15835","kind":"arxiv","version":1}},"canonical_sha256":"4702427ebe69bd3e99d7a64ae31bab92d1210b41c9d3a44ff842bf5bceb13c8b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4702427ebe69bd3e99d7a64ae31bab92d1210b41c9d3a44ff842bf5bceb13c8b","first_computed_at":"2026-07-05T11:57:21.732254Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:57:21.732254Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"jbTdQzno5KdQWEcdn59MiMvTBT04Xbl4sg0jxCmGBQOh/Xx+rSAzhy9twwz8EVAhD29Wj6CbTDF0wLtsRocsDg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:57:21.732692Z","signed_message":"canonical_sha256_bytes"},"source_id":"2508.15835","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:14675c3774052f58321e2800e064f3a7bf79f9b0f9bfaf46e829ba2cedf71409","sha256:6ebc98b10a1bd776f6cc60d753e5c8430a65cdbddaef7c6f0ba1fc3935207e8b"],"state_sha256":"da090961071a3007290b4e6645cb4e2e041bf61095ecf4e40d4f9826c68f8860"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Yhfq6KyoAKKhCQ2231iT33KTAWxWXB+CAP9Jmu7Muk6zSv2a/neOIBbS1uJlF2jwQgkrZ8Y8C7U+DHwuRkOwCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T03:05:40.503007Z","bundle_sha256":"337593135521e69bdf076e69e8e2cc557fd9c30871829fa5205a70a86fce5544"}}