{"paper":{"title":"Sentinel-VLA: A Metacognitive VLA Model with Active Status Monitoring for Dynamic Reasoning and Error Recovery","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Sentinel-VLA equips VLA models with an active sentinel module that monitors execution status and triggers reasoning or error recovery only when needed, delivering over 30% higher real-world task success.","cross_cats":[],"primary_cat":"cs.RO","authors_text":"Chang Xu, Dan Niu, Hongyan Xu, Lei Fan, Shan You, Wenhao Li, Xiu Su, Yichao Cao, Zhe Qu","submitted_at":"2026-05-02T02:10:54Z","abstract_excerpt":"Vision-language-action (VLA) models have advanced the field of embodied manipulation by harnessing broad world knowledge and strong generalization. However, current VLA models still face several key challenges, including limited reasoning capability, lack of status monitoring, and difficulty in self-correction. In this paper, we introduce \\textbf{Sentinel-VLA}, a metacognitive VLA model equipped with an active ``sentinel'' module to monitor real-time execution status. Only when necessary, such as during initial planning or upon detecting an error, the model triggers a dynamic reasoning or form"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Real-world experiments demonstrate that Sentinel-VLA boosts the task success rate by over 30% compared to the SOTA model, PI0.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The automatically generated and annotated training data (44 tasks, 2.6 million transitions) faithfully captures the distribution of real-world execution errors and recovery opportunities without introducing systematic biases from the generation pipeline.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Sentinel-VLA introduces a metacognitive VLA model with a sentinel module for real-time status monitoring, dynamic reasoning, and error recovery, plus a self-evolving continual learning method, raising real-world task success by over 30% versus prior SOTA.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Sentinel-VLA equips VLA models with an active sentinel module that monitors execution status and triggers reasoning or error recovery only when needed, delivering over 30% higher real-world task success.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"2e31368f094d2cbfb8db9227b770211fd9580e347046aa14f13ac9d17173b5fb"},"source":{"id":"2605.01191","kind":"arxiv","version":2},"verdict":{"id":"14ae98ab-57d5-4377-8a7b-e1586bd4986c","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-09T15:10:44.352890Z","strongest_claim":"Real-world experiments demonstrate that Sentinel-VLA boosts the task success rate by over 30% compared to the SOTA model, PI0.","one_line_summary":"Sentinel-VLA introduces a metacognitive VLA model with a sentinel module for real-time status monitoring, dynamic reasoning, and error recovery, plus a self-evolving continual learning method, raising real-world task success by over 30% versus prior SOTA.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The automatically generated and annotated training data (44 tasks, 2.6 million transitions) faithfully captures the distribution of real-world execution errors and recovery opportunities without introducing systematic biases from the generation pipeline.","pith_extraction_headline":"Sentinel-VLA equips VLA models with an active sentinel module that monitors execution status and triggers reasoning or error recovery only when needed, delivering over 30% higher real-world task success."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.01191/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-20T18:36:57.526473Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T17:30:27.034694Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"64730db73bece1b644444dee29729281a6f10fcb616df4f648f383674125ab99"},"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"}