{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:BIK2UAXAY6QUYII2MKAKHEQUDT","short_pith_number":"pith:BIK2UAXA","schema_version":"1.0","canonical_sha256":"0a15aa02e0c7a14c211a6280a392141cfc00f90a52c2bd02063678c87bc3dd37","source":{"kind":"arxiv","id":"2605.20774","version":1},"attestation_state":"computed","paper":{"title":"VLA-REPLICA: A Low-Cost, Reproducible Benchmark for Real-World Evaluation of Vision-Language-Action Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Alex S. Huang, Jiahui Zhang, Shiqing Tang, Yu Xiang","submitted_at":"2026-05-20T06:15:30Z","abstract_excerpt":"Vision-Language-Action (VLA) models have shown strong promise for general-purpose robotic manipulation, but their real-world evaluation remains limited by a lack of accessible, reproducible, and consistent benchmarks. Simulation benchmarks fail to capture real-world complexity, while existing real-world benchmarks often require expensive hardware, centralized evaluation, or are limited in task diversity. We introduce VLA-REPLICA, a low-cost, easily reproducible real-world benchmark for evaluating VLA models. Built from off-the-shelf components, our system can be quickly assembled and replicate"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2605.20774","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2026-05-20T06:15:30Z","cross_cats_sorted":[],"title_canon_sha256":"78094864e7de384ed9444640c8bced5ec07654033bb9fe97cd3073ec5783c169","abstract_canon_sha256":"814a599d71576c03dccd4117fa85892bd826842f08ba6e20390cdae7cdc204ea"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:04:53.609943Z","signature_b64":"PLLq6swG8I/RK9sY+kdFGjtCNEDyi+RbkPVAyKu2fAav0eeJSfwkKw8mL4q8idv9Th9adnE8dZuJ/go0doBkCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0a15aa02e0c7a14c211a6280a392141cfc00f90a52c2bd02063678c87bc3dd37","last_reissued_at":"2026-05-21T01:04:53.609183Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:04:53.609183Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"VLA-REPLICA: A Low-Cost, Reproducible Benchmark for Real-World Evaluation of Vision-Language-Action Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Alex S. Huang, Jiahui Zhang, Shiqing Tang, Yu Xiang","submitted_at":"2026-05-20T06:15:30Z","abstract_excerpt":"Vision-Language-Action (VLA) models have shown strong promise for general-purpose robotic manipulation, but their real-world evaluation remains limited by a lack of accessible, reproducible, and consistent benchmarks. Simulation benchmarks fail to capture real-world complexity, while existing real-world benchmarks often require expensive hardware, centralized evaluation, or are limited in task diversity. We introduce VLA-REPLICA, a low-cost, easily reproducible real-world benchmark for evaluating VLA models. Built from off-the-shelf components, our system can be quickly assembled and replicate"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.20774","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/2605.20774/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2605.20774","created_at":"2026-05-21T01:04:53.609329+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.20774v1","created_at":"2026-05-21T01:04:53.609329+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.20774","created_at":"2026-05-21T01:04:53.609329+00:00"},{"alias_kind":"pith_short_12","alias_value":"BIK2UAXAY6QU","created_at":"2026-05-21T01:04:53.609329+00:00"},{"alias_kind":"pith_short_16","alias_value":"BIK2UAXAY6QUYII2","created_at":"2026-05-21T01:04:53.609329+00:00"},{"alias_kind":"pith_short_8","alias_value":"BIK2UAXA","created_at":"2026-05-21T01:04:53.609329+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BIK2UAXAY6QUYII2MKAKHEQUDT","json":"https://pith.science/pith/BIK2UAXAY6QUYII2MKAKHEQUDT.json","graph_json":"https://pith.science/api/pith-number/BIK2UAXAY6QUYII2MKAKHEQUDT/graph.json","events_json":"https://pith.science/api/pith-number/BIK2UAXAY6QUYII2MKAKHEQUDT/events.json","paper":"https://pith.science/paper/BIK2UAXA"},"agent_actions":{"view_html":"https://pith.science/pith/BIK2UAXAY6QUYII2MKAKHEQUDT","download_json":"https://pith.science/pith/BIK2UAXAY6QUYII2MKAKHEQUDT.json","view_paper":"https://pith.science/paper/BIK2UAXA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.20774&json=true","fetch_graph":"https://pith.science/api/pith-number/BIK2UAXAY6QUYII2MKAKHEQUDT/graph.json","fetch_events":"https://pith.science/api/pith-number/BIK2UAXAY6QUYII2MKAKHEQUDT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BIK2UAXAY6QUYII2MKAKHEQUDT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BIK2UAXAY6QUYII2MKAKHEQUDT/action/storage_attestation","attest_author":"https://pith.science/pith/BIK2UAXAY6QUYII2MKAKHEQUDT/action/author_attestation","sign_citation":"https://pith.science/pith/BIK2UAXAY6QUYII2MKAKHEQUDT/action/citation_signature","submit_replication":"https://pith.science/pith/BIK2UAXAY6QUYII2MKAKHEQUDT/action/replication_record"}},"created_at":"2026-05-21T01:04:53.609329+00:00","updated_at":"2026-05-21T01:04:53.609329+00:00"}