{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:MPAPB77ZT35B2NCP5J4R3ZIVAS","short_pith_number":"pith:MPAPB77Z","canonical_record":{"source":{"id":"2505.12844","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-05-19T08:30:13Z","cross_cats_sorted":["cs.RO"],"title_canon_sha256":"fe79aaaf291b3b55df84fb1d50e7f7396e6af4619e550a3891d88402e1032dbf","abstract_canon_sha256":"1769710b0fbae38805830c6998196151560dc4ad2ac043a2df22c46bd0c73ba3"},"schema_version":"1.0"},"canonical_sha256":"63c0f0fff99efa1d344fea791de51504a285ff9323f7d9757e32f829a4522862","source":{"kind":"arxiv","id":"2505.12844","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.12844","created_at":"2026-07-05T11:08:41Z"},{"alias_kind":"arxiv_version","alias_value":"2505.12844v2","created_at":"2026-07-05T11:08:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12844","created_at":"2026-07-05T11:08:41Z"},{"alias_kind":"pith_short_12","alias_value":"MPAPB77ZT35B","created_at":"2026-07-05T11:08:41Z"},{"alias_kind":"pith_short_16","alias_value":"MPAPB77ZT35B2NCP","created_at":"2026-07-05T11:08:41Z"},{"alias_kind":"pith_short_8","alias_value":"MPAPB77Z","created_at":"2026-07-05T11:08:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:MPAPB77ZT35B2NCP5J4R3ZIVAS","target":"record","payload":{"canonical_record":{"source":{"id":"2505.12844","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-05-19T08:30:13Z","cross_cats_sorted":["cs.RO"],"title_canon_sha256":"fe79aaaf291b3b55df84fb1d50e7f7396e6af4619e550a3891d88402e1032dbf","abstract_canon_sha256":"1769710b0fbae38805830c6998196151560dc4ad2ac043a2df22c46bd0c73ba3"},"schema_version":"1.0"},"canonical_sha256":"63c0f0fff99efa1d344fea791de51504a285ff9323f7d9757e32f829a4522862","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:08:41.444647Z","signature_b64":"81Q6QHicoEkarrEdXPljuTaNjzPD3fjW6b3JQNvgg4xOav6Cwv7kXdC9JyvV/b93rP0mhT1woFc1NSTABhAyAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"63c0f0fff99efa1d344fea791de51504a285ff9323f7d9757e32f829a4522862","last_reissued_at":"2026-07-05T11:08:41.444211Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:08:41.444211Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.12844","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:08:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2/Z93nxm5L9bWbT2IUKvu9BRRPL0pKweu4bZwy5V2HLPB8f3OfTST+CMhHW3upRr3srHZ88V0JQuycp1RMXRAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T07:37:23.435484Z"},"content_sha256":"b0d81e6f710b9a3d63709aae02618d2b968d872797da6532f54404736fcac51d","schema_version":"1.0","event_id":"sha256:b0d81e6f710b9a3d63709aae02618d2b968d872797da6532f54404736fcac51d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:MPAPB77ZT35B2NCP5J4R3ZIVAS","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"AGI-Elo: How Far Are We From Mastering A Task?","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.RO"],"primary_cat":"cs.AI","authors_text":"Alok Prakash, Chengran Yuan, Christina Dao Wen Lee, Dongen Li, Jiawei Sun, Justin KW Yeoh, Marcelo H. Ang Jr, Shuo Sun, Thomas W. Malone, Yimin Zhao, Zefan Huang","submitted_at":"2025-05-19T08:30:13Z","abstract_excerpt":"As the field progresses toward Artificial General Intelligence (AGI), there is a pressing need for more comprehensive and insightful evaluation frameworks that go beyond aggregate performance metrics. This paper introduces a unified rating system that jointly models the difficulty of individual test cases and the competency of AI models (or humans) across vision, language, and action domains. Unlike existing metrics that focus solely on models, our approach allows for fine-grained, difficulty-aware evaluations through competitive interactions between models and tasks, capturing both the long-t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12844","kind":"arxiv","version":2},"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/2505.12844/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":"2026-07-05T11:08:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mtSGk1ee5ioHapxSBJqQ8Sh0vmv1EQPKMDKAN6fmdQ45Kv4o44b9nX4k/HPY6G1rfxtMoE3JQOCLf9Cv4g+FDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-16T07:37:23.436418Z"},"content_sha256":"dfff243002eba557d52978b527d109bdb8c065f62fb5163d969b6d89a852b342","schema_version":"1.0","event_id":"sha256:dfff243002eba557d52978b527d109bdb8c065f62fb5163d969b6d89a852b342"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MPAPB77ZT35B2NCP5J4R3ZIVAS/bundle.json","state_url":"https://pith.science/pith/MPAPB77ZT35B2NCP5J4R3ZIVAS/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MPAPB77ZT35B2NCP5J4R3ZIVAS/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":"2026-08-16T07:37:23Z","links":{"resolver":"https://pith.science/pith/MPAPB77ZT35B2NCP5J4R3ZIVAS","bundle":"https://pith.science/pith/MPAPB77ZT35B2NCP5J4R3ZIVAS/bundle.json","state":"https://pith.science/pith/MPAPB77ZT35B2NCP5J4R3ZIVAS/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MPAPB77ZT35B2NCP5J4R3ZIVAS/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:MPAPB77ZT35B2NCP5J4R3ZIVAS","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":"1769710b0fbae38805830c6998196151560dc4ad2ac043a2df22c46bd0c73ba3","cross_cats_sorted":["cs.RO"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-05-19T08:30:13Z","title_canon_sha256":"fe79aaaf291b3b55df84fb1d50e7f7396e6af4619e550a3891d88402e1032dbf"},"schema_version":"1.0","source":{"id":"2505.12844","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.12844","created_at":"2026-07-05T11:08:41Z"},{"alias_kind":"arxiv_version","alias_value":"2505.12844v2","created_at":"2026-07-05T11:08:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12844","created_at":"2026-07-05T11:08:41Z"},{"alias_kind":"pith_short_12","alias_value":"MPAPB77ZT35B","created_at":"2026-07-05T11:08:41Z"},{"alias_kind":"pith_short_16","alias_value":"MPAPB77ZT35B2NCP","created_at":"2026-07-05T11:08:41Z"},{"alias_kind":"pith_short_8","alias_value":"MPAPB77Z","created_at":"2026-07-05T11:08:41Z"}],"graph_snapshots":[{"event_id":"sha256:dfff243002eba557d52978b527d109bdb8c065f62fb5163d969b6d89a852b342","target":"graph","created_at":"2026-07-05T11:08:41Z","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/2505.12844/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As the field progresses toward Artificial General Intelligence (AGI), there is a pressing need for more comprehensive and insightful evaluation frameworks that go beyond aggregate performance metrics. This paper introduces a unified rating system that jointly models the difficulty of individual test cases and the competency of AI models (or humans) across vision, language, and action domains. Unlike existing metrics that focus solely on models, our approach allows for fine-grained, difficulty-aware evaluations through competitive interactions between models and tasks, capturing both the long-t","authors_text":"Alok Prakash, Chengran Yuan, Christina Dao Wen Lee, Dongen Li, Jiawei Sun, Justin KW Yeoh, Marcelo H. Ang Jr, Shuo Sun, Thomas W. Malone, Yimin Zhao, Zefan Huang","cross_cats":["cs.RO"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-05-19T08:30:13Z","title":"AGI-Elo: How Far Are We From Mastering A Task?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12844","kind":"arxiv","version":2},"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:b0d81e6f710b9a3d63709aae02618d2b968d872797da6532f54404736fcac51d","target":"record","created_at":"2026-07-05T11:08:41Z","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":"1769710b0fbae38805830c6998196151560dc4ad2ac043a2df22c46bd0c73ba3","cross_cats_sorted":["cs.RO"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-05-19T08:30:13Z","title_canon_sha256":"fe79aaaf291b3b55df84fb1d50e7f7396e6af4619e550a3891d88402e1032dbf"},"schema_version":"1.0","source":{"id":"2505.12844","kind":"arxiv","version":2}},"canonical_sha256":"63c0f0fff99efa1d344fea791de51504a285ff9323f7d9757e32f829a4522862","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"63c0f0fff99efa1d344fea791de51504a285ff9323f7d9757e32f829a4522862","first_computed_at":"2026-07-05T11:08:41.444211Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:08:41.444211Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"81Q6QHicoEkarrEdXPljuTaNjzPD3fjW6b3JQNvgg4xOav6Cwv7kXdC9JyvV/b93rP0mhT1woFc1NSTABhAyAg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:08:41.444647Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.12844","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b0d81e6f710b9a3d63709aae02618d2b968d872797da6532f54404736fcac51d","sha256:dfff243002eba557d52978b527d109bdb8c065f62fb5163d969b6d89a852b342"],"state_sha256":"2eb38bfc141afe89b64f568d51d8965c08cc1240993653efd61f9325a3cbc172"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"kauCffEejOc57YrkFEk0IkpdMLRwQxIMRlHQY1gTX22oW30MDzVIPdK8G+a3n5bMJv/ugifCtpA1Uu29fWLZBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-16T07:37:23.509822Z","bundle_sha256":"6da83e35a01ecc6977ac8bf60fdd927b0861120cc4b44bdd858f11a4f20ef6d3"}}