{"paper":{"title":"Focal plane wavefront control with model-based reinforcement learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A model-based reinforcement learning algorithm corrects non-common-path aberrations from focal-plane images alone.","cross_cats":["astro-ph.EP","cs.LG","cs.RO"],"primary_cat":"astro-ph.IM","authors_text":"Gilles Orban De Xivry, Iremsu Taskin, Jalo Nousiainen, Markus Kasper, Olivier Absil","submitted_at":"2026-04-01T14:55:15Z","abstract_excerpt":"The direct imaging of potentially habitable exoplanets is one prime science case for high-contrast imaging instruments on extremely large telescopes. Most such exoplanets orbit close to their host stars, where their observation is limited by fast-moving atmospheric speckles and quasi-static non-common-path aberrations (NCPA). Conventional NCPA correction methods often use mechanical mirror probes, which compromise performance during operation. This work presents machine-learning-based NCPA control methods that automatically detect and correct both dynamic and static NCPA errors by leveraging s"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Simulations show that PO4NCPA robustly compensates static and dynamic NCPAs. In static cases, it achieves near-optimal focal-plane light suppression with a coronagraph and near-optimal Strehl without one. With dynamic NCPA, it matches the performance of the modal least-squares reconstruction combined with a 1-step delay integrator.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The numerical simulations of static NCPA on a ground-based telescope and water-vapor-induced dynamic NCPA accurately capture the statistics, noise properties, and temporal behavior that will be encountered on real ELT-class instruments with a vector vortex coronagraph.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"PO4NCPA applies model-based RL to focal-plane images via sequential phase diversity to correct static and dynamic NCPAs, achieving near-optimal suppression in simulations for coronagraphic and non-coronagraphic cases.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A model-based reinforcement learning algorithm corrects non-common-path aberrations from focal-plane images alone.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"689fb9f07869769808a3ff8986bd55f59877db5cf74a4acdfb961c3ecb12aae7"},"source":{"id":"2604.00993","kind":"arxiv","version":1},"verdict":{"id":"ed03cdf8-b730-41f2-87e9-bfaba7fca193","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-13T22:00:03.951097Z","strongest_claim":"Simulations show that PO4NCPA robustly compensates static and dynamic NCPAs. In static cases, it achieves near-optimal focal-plane light suppression with a coronagraph and near-optimal Strehl without one. With dynamic NCPA, it matches the performance of the modal least-squares reconstruction combined with a 1-step delay integrator.","one_line_summary":"PO4NCPA applies model-based RL to focal-plane images via sequential phase diversity to correct static and dynamic NCPAs, achieving near-optimal suppression in simulations for coronagraphic and non-coronagraphic cases.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The numerical simulations of static NCPA on a ground-based telescope and water-vapor-induced dynamic NCPA accurately capture the statistics, noise properties, and temporal behavior that will be encountered on real ELT-class instruments with a vector vortex coronagraph.","pith_extraction_headline":"A model-based reinforcement learning algorithm corrects non-common-path aberrations from focal-plane images alone."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.00993/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":54,"sample":[{"doi":"","year":2022,"title":"Absil, O., Delacroix, C., Orban de Xivry, G., et al. 2022, in Proc. SPIE Conf., V ol. 12185, SPIE, 298–310","work_id":"aa0a4bc6-3ca2-40e7-892a-c718bd76cc02","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2016,"title":"Absil, O., Mawet, D., Karlsson, M., et al. 2016, in Proc. SPIE Conf., V ol. 9908, 99080Q","work_id":"2a3fa8fa-c464-4e81-a2f6-5833d53058ba","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":1990,"title":"Angel, J. R. P., Wizinowich, P., Lloyd-Hart, M., & Sandler, D. 1990, Nat, 348, 221","work_id":"95d7d506-aad5-4c96-9421-4f9c0f466405","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2020,"title":"P., Vievard, S., Wilby, M","work_id":"1d46a214-9be4-4a66-8f1b-ad9968c36897","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"Bottom, M., Walker, S. A. U., Cunnyngham, I., Guthery, C., & Delorme, J.-R. 2023, arXiv e-prints, arXiv:2312.06806","work_id":"0e8bda5f-26ae-486a-9a5e-e8cff701792a","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":54,"snapshot_sha256":"8c8eb68a9a99d5994033db58719401fda29dc5bbfa9d5dfb9c0ab1bb04b40a53","internal_anchors":1},"formal_canon":{"evidence_count":1,"snapshot_sha256":"29f3c0827c5a66517dbc73fa59e8f6b8157b771cf791ec8e89f88af9ff1fb498"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}