Predictive Physical AI systems output state rollouts, action chunks, and latent plans, yet a low root-mean-square error (RMSE) does not imply that a particular proposal is physically executable. We formulate physical admissibility as a prediction-control interface: before execution, a decoded proposal is treated as candidate dynamics and evaluated using kinematic, dynamic, and direct-to-composed horizon conditions. Passing is not a certificate of task success; rejection identifies violation of the specified physical envelope and gives a component-level reason. On Hugging Face LeRobot PushT, controlled falsification shows that one-step prediction-RMSE and standardized dynamics residuals reach area under the receiver operating characteristic curve (AUC) 0.982 and 0.972, kinematic-only conditions reach AUC 0.592, and the full gate reaches AUC 0.957 with condition-level attribution. In replay-based intervention experiments, residual-based filters and the full physical-admissibility gate prevent 87-$89% of invalid proposals while preserving mean progress near 0.998.</p>\n","updatedAt":"2026-06-02T07:26:05.935Z","author":{"_id":"6a05ebd6a01745697ea77cc0","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6a05ebd6a01745697ea77cc0/GCFxXC7FSR8inb7kSJ8Bc.jpeg","fullname":"Barak Or","name":"barakor","type":"user","isPro":true,"isHf":false,"isHfAdmin":false,"isMod":false,"isUserFollowing":false}},"numEdits":0,"identifiedLanguage":{"language":"en","probability":0.878499448299408},"editors":["barakor"],"editorAvatarUrls":["https://cdn-avatars.huggingface.co/v1/production/uploads/6a05ebd6a01745697ea77cc0/GCFxXC7FSR8inb7kSJ8Bc.jpeg"],"reactions":[],"isReport":false}}],"primaryEmailConfirmed":false,"paper":{"id":"2606.00089","authors":[{"_id":"6a1e8526808ddbc3c7d43f33","name":"Barak Or","hidden":false}],"publishedAt":"2026-05-23T00:00:00.000Z","submittedOnDailyAt":"2026-06-02T00:00:00.000Z","title":"Can Predicted Dynamics Exist in the Physical World?","submittedOnDailyBy":{"_id":"6a05ebd6a01745697ea77cc0","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6a05ebd6a01745697ea77cc0/GCFxXC7FSR8inb7kSJ8Bc.jpeg","isPro":true,"fullname":"Barak Or","user":"barakor","type":"user","name":"barakor"},"summary":"Predictive Physical AI systems output state rollouts, action chunks, and latent plans, yet a low root-mean-square error (RMSE) does not imply that a particular proposal is physically executable. We formulate physical admissibility as a prediction-control interface: before execution, a decoded proposal is treated as candidate dynamics and evaluated using kinematic, dynamic, and direct-to-composed horizon conditions. Passing is not a certificate of task success; rejection identifies violation of the specified physical envelope and gives a component-level reason. On Hugging Face LeRobot PushT, controlled falsification shows that one-step prediction-RMSE and standardized dynamics residuals reach area under the receiver operating characteristic curve (AUC) 0.982 and 0.972, kinematic-only conditions reach AUC 0.592, and the full gate reaches AUC 0.957 with condition-level attribution. In replay-based intervention experiments, residual-based filters and the full physical-admissibility gate prevent 87-$89% of invalid proposals while preserving mean progress near 0.998.","upvotes":2,"discussionId":"6a1e8526808ddbc3c7d43f34","projectPage":"http://www.state16.ai","ai_summary":"Physical admissibility validation for AI systems uses prediction-control interfaces with kinematic and dynamic conditions to filter invalid proposals while maintaining high performance.","ai_keywords":["predictive physical AI","state rollouts","action chunks","latent plans","root-mean-square error","prediction-control interface","kinematic conditions","dynamic conditions","direct-to-composed horizon conditions","physical admissibility","receiver operating characteristic curve","AUC","residual-based filters"],"organization":{"_id":"6a05ec439bca8dbc65d0aef6","name":"STATE16","fullname":"STATE16","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/6a05ebd6a01745697ea77cc0/K2Y4PVuv5MAy14uwy3R8y.jpeg"}},"canReadDatabase":false,"canManagePapers":false,"canSubmit":false,"hasHfLevelAccess":false,"upvoted":false,"upvoters":[{"_id":"6a05ebd6a01745697ea77cc0","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/6a05ebd6a01745697ea77cc0/GCFxXC7FSR8inb7kSJ8Bc.jpeg","isPro":true,"fullname":"Barak Or","user":"barakor","type":"user"},{"_id":"69bd57cb0b88f54e45f721a2","avatarUrl":"https://cdn-avatars.huggingface.co/v1/production/uploads/noauth/WiOsERDLnvRFc1VfI4-fw.png","isPro":false,"fullname":"Ezra Green","user":"ajones2025","type":"user"}],"acceptLanguages":["en"],"dailyPaperRank":0,"organization":{"_id":"6a05ec439bca8dbc65d0aef6","name":"STATE16","fullname":"STATE16","avatar":"https://cdn-avatars.huggingface.co/v1/production/uploads/6a05ebd6a01745697ea77cc0/K2Y4PVuv5MAy14uwy3R8y.jpeg"},"markdownContentUrl":"https://huggingface.co/buckets/huggingchat/papers-content/resolve/2606/2606.00089.md"}">
Can Predicted Dynamics Exist in the Physical World?
Abstract
Physical admissibility validation for AI systems uses prediction-control interfaces with kinematic and dynamic conditions to filter invalid proposals while maintaining high performance.
AI-generated summary
Predictive Physical AI systems output state rollouts, action chunks, and latent plans, yet a low root-mean-square error (RMSE) does not imply that a particular proposal is physically executable. We formulate physical admissibility as a prediction-control interface: before execution, a decoded proposal is treated as candidate dynamics and evaluated using kinematic, dynamic, and direct-to-composed horizon conditions. Passing is not a certificate of task success; rejection identifies violation of the specified physical envelope and gives a component-level reason. On Hugging Face LeRobot PushT, controlled falsification shows that one-step prediction-RMSE and standardized dynamics residuals reach area under the receiver operating characteristic curve (AUC) 0.982 and 0.972, kinematic-only conditions reach AUC 0.592, and the full gate reaches AUC 0.957 with condition-level attribution. In replay-based intervention experiments, residual-based filters and the full physical-admissibility gate prevent 87-$89% of invalid proposals while preserving mean progress near 0.998.
Community
Predictive Physical AI systems output state rollouts, action chunks, and latent plans, yet a low root-mean-square error (RMSE) does not imply that a particular proposal is physically executable. We formulate physical admissibility as a prediction-control interface: before execution, a decoded proposal is treated as candidate dynamics and evaluated using kinematic, dynamic, and direct-to-composed horizon conditions. Passing is not a certificate of task success; rejection identifies violation of the specified physical envelope and gives a component-level reason. On Hugging Face LeRobot PushT, controlled falsification shows that one-step prediction-RMSE and standardized dynamics residuals reach area under the receiver operating characteristic curve (AUC) 0.982 and 0.972, kinematic-only conditions reach AUC 0.592, and the full gate reaches AUC 0.957 with condition-level attribution. In replay-based intervention experiments, residual-based filters and the full physical-admissibility gate prevent 87-$89% of invalid proposals while preserving mean progress near 0.998.
Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.
Tap or paste here to upload images
Cite arxiv.org/abs/2606.00089 in a model README.md to link it from this page.
Cite arxiv.org/abs/2606.00089 in a dataset README.md to link it from this page.
Cite arxiv.org/abs/2606.00089 in a Space README.md to link it from this page.
Discussion (0)
Sign in to join the discussion. Free account, 30 seconds — email code or GitHub.
Sign in →No comments yet. Sign in and be the first to say something.