{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:5A3LG6JDWOPITIWBCFX6B2OK3L","short_pith_number":"pith:5A3LG6JD","canonical_record":{"source":{"id":"2503.24278","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2025-03-31T16:23:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c048c137a2c47d639fd1ac4a14cd030864980fdd1204346f7cf19b3f4f460a3c","abstract_canon_sha256":"2907f9b45a7c043a7c8f947aa85c6a6a4c4719480fa3d1d397a27c671df16176"},"schema_version":"1.0"},"canonical_sha256":"e836b37923b39e89a2c1116fe0e9cadaf7809220d8a593e6f65496c9797b750d","source":{"kind":"arxiv","id":"2503.24278","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.24278","created_at":"2026-07-05T10:43:47Z"},{"alias_kind":"arxiv_version","alias_value":"2503.24278v2","created_at":"2026-07-05T10:43:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.24278","created_at":"2026-07-05T10:43:47Z"},{"alias_kind":"pith_short_12","alias_value":"5A3LG6JDWOPI","created_at":"2026-07-05T10:43:47Z"},{"alias_kind":"pith_short_16","alias_value":"5A3LG6JDWOPITIWB","created_at":"2026-07-05T10:43:47Z"},{"alias_kind":"pith_short_8","alias_value":"5A3LG6JD","created_at":"2026-07-05T10:43:47Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:5A3LG6JDWOPITIWBCFX6B2OK3L","target":"record","payload":{"canonical_record":{"source":{"id":"2503.24278","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2025-03-31T16:23:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c048c137a2c47d639fd1ac4a14cd030864980fdd1204346f7cf19b3f4f460a3c","abstract_canon_sha256":"2907f9b45a7c043a7c8f947aa85c6a6a4c4719480fa3d1d397a27c671df16176"},"schema_version":"1.0"},"canonical_sha256":"e836b37923b39e89a2c1116fe0e9cadaf7809220d8a593e6f65496c9797b750d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:43:47.045023Z","signature_b64":"8tV9BPsxm35UHePUCmTF7lCQeYJV4Xqf6Vf5pw76EJ2tn1eWu3Tq8z3qwbwbBgvVQpFqSXLCltF8mosNskTAAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e836b37923b39e89a2c1116fe0e9cadaf7809220d8a593e6f65496c9797b750d","last_reissued_at":"2026-07-05T10:43:47.044565Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:43:47.044565Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.24278","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":"2026-07-05T10:43:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ed+5CO3kR5slyev62GBIw/d2SymGtVkicYwKL9tR8ZRoLIRMYi0XNqkGMRC+A8CYpsUMBh9206XVpYGLBprRCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T04:58:05.806311Z"},"content_sha256":"68472d8bead6ea18b79bb83e4fa928f7dbbaaf083a8737cf3651e4dee844d250","schema_version":"1.0","event_id":"sha256:68472d8bead6ea18b79bb83e4fa928f7dbbaaf083a8737cf3651e4dee844d250"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:5A3LG6JDWOPITIWBCFX6B2OK3L","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"AutoEval: Autonomous Evaluation of Generalist Robot Manipulation Policies in the Real World","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.RO","authors_text":"Karl Pertsch, Pranav Atreya, Sergey Levine, You Liang Tan, Zhiyuan Zhou","submitted_at":"2025-03-31T16:23:44Z","abstract_excerpt":"Scalable and reproducible policy evaluation has been a long-standing challenge in robot learning. Evaluations are critical to assess progress and build better policies, but evaluation in the real world, especially at a scale that would provide statistically reliable results, is costly in terms of human time and hard to obtain. Evaluation of increasingly generalist robot policies requires an increasingly diverse repertoire of evaluation environments, making the evaluation bottleneck even more pronounced. To make real-world evaluation of robotic policies more practical, we propose AutoEval, a sy"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.24278","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/2503.24278/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":"2026-07-05T10:43:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qLI4OsOFZlcKFPF+CC6ZqHuaCa5KW/Ho3OXUGlOUir+/2h+/Cs9TMrjdEFZW49W3MqqROAhxQlarK82U1yr6AA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T04:58:05.807259Z"},"content_sha256":"db4b19ec101c2cb34f3be19d723c22e44913e10038aa55dd61dcea7d7b7a4216","schema_version":"1.0","event_id":"sha256:db4b19ec101c2cb34f3be19d723c22e44913e10038aa55dd61dcea7d7b7a4216"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/5A3LG6JDWOPITIWBCFX6B2OK3L/bundle.json","state_url":"https://pith.science/pith/5A3LG6JDWOPITIWBCFX6B2OK3L/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/5A3LG6JDWOPITIWBCFX6B2OK3L/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":"2026-08-05T04:58:05Z","links":{"resolver":"https://pith.science/pith/5A3LG6JDWOPITIWBCFX6B2OK3L","bundle":"https://pith.science/pith/5A3LG6JDWOPITIWBCFX6B2OK3L/bundle.json","state":"https://pith.science/pith/5A3LG6JDWOPITIWBCFX6B2OK3L/state.json","well_known_bundle":"https://pith.science/.well-known/pith/5A3LG6JDWOPITIWBCFX6B2OK3L/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:5A3LG6JDWOPITIWBCFX6B2OK3L","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":"2907f9b45a7c043a7c8f947aa85c6a6a4c4719480fa3d1d397a27c671df16176","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2025-03-31T16:23:44Z","title_canon_sha256":"c048c137a2c47d639fd1ac4a14cd030864980fdd1204346f7cf19b3f4f460a3c"},"schema_version":"1.0","source":{"id":"2503.24278","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.24278","created_at":"2026-07-05T10:43:47Z"},{"alias_kind":"arxiv_version","alias_value":"2503.24278v2","created_at":"2026-07-05T10:43:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.24278","created_at":"2026-07-05T10:43:47Z"},{"alias_kind":"pith_short_12","alias_value":"5A3LG6JDWOPI","created_at":"2026-07-05T10:43:47Z"},{"alias_kind":"pith_short_16","alias_value":"5A3LG6JDWOPITIWB","created_at":"2026-07-05T10:43:47Z"},{"alias_kind":"pith_short_8","alias_value":"5A3LG6JD","created_at":"2026-07-05T10:43:47Z"}],"graph_snapshots":[{"event_id":"sha256:db4b19ec101c2cb34f3be19d723c22e44913e10038aa55dd61dcea7d7b7a4216","target":"graph","created_at":"2026-07-05T10:43:47Z","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/2503.24278/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Scalable and reproducible policy evaluation has been a long-standing challenge in robot learning. Evaluations are critical to assess progress and build better policies, but evaluation in the real world, especially at a scale that would provide statistically reliable results, is costly in terms of human time and hard to obtain. Evaluation of increasingly generalist robot policies requires an increasingly diverse repertoire of evaluation environments, making the evaluation bottleneck even more pronounced. To make real-world evaluation of robotic policies more practical, we propose AutoEval, a sy","authors_text":"Karl Pertsch, Pranav Atreya, Sergey Levine, You Liang Tan, Zhiyuan Zhou","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2025-03-31T16:23:44Z","title":"AutoEval: Autonomous Evaluation of Generalist Robot Manipulation Policies in the Real World"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.24278","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":"sha256:68472d8bead6ea18b79bb83e4fa928f7dbbaaf083a8737cf3651e4dee844d250","target":"record","created_at":"2026-07-05T10:43:47Z","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":"2907f9b45a7c043a7c8f947aa85c6a6a4c4719480fa3d1d397a27c671df16176","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2025-03-31T16:23:44Z","title_canon_sha256":"c048c137a2c47d639fd1ac4a14cd030864980fdd1204346f7cf19b3f4f460a3c"},"schema_version":"1.0","source":{"id":"2503.24278","kind":"arxiv","version":2}},"canonical_sha256":"e836b37923b39e89a2c1116fe0e9cadaf7809220d8a593e6f65496c9797b750d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e836b37923b39e89a2c1116fe0e9cadaf7809220d8a593e6f65496c9797b750d","first_computed_at":"2026-07-05T10:43:47.044565Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:43:47.044565Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8tV9BPsxm35UHePUCmTF7lCQeYJV4Xqf6Vf5pw76EJ2tn1eWu3Tq8z3qwbwbBgvVQpFqSXLCltF8mosNskTAAA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:43:47.045023Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.24278","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:68472d8bead6ea18b79bb83e4fa928f7dbbaaf083a8737cf3651e4dee844d250","sha256:db4b19ec101c2cb34f3be19d723c22e44913e10038aa55dd61dcea7d7b7a4216"],"state_sha256":"07fdb4303dbe01b1cd370ee8f336b349943b3e91036ce86b788f2f784dd0f387"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mww3Q9eHdfcaKsD+iglHNl0fH7Gazo/Yu5HMZUgNWg74o7y1S7GFEEX55GatcVnivN/bLJ4g2aOhoRzIyfSaDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T04:58:05.814733Z","bundle_sha256":"d3c89c3690e4d0cfc6b8f2f77af85af97bf14e81773ff0b47097f94af7f9e984"}}