| { |
| "@context": { |
| "@language": "en", |
| "@vocab": "https://schema.org/", |
| "conformsTo": "dct:conformsTo", |
| "cr": "http://mlcommons.org/croissant/", |
| "dataType": { |
| "@id": "cr:dataType", |
| "@type": "@vocab" |
| }, |
| "dct": "http://purl.org/dc/terms/", |
| "prov": "http://www.w3.org/ns/prov#", |
| "rai": "http://mlcommons.org/croissant/RAI/", |
| "sc": "https://schema.org/", |
| "value": { |
| "@id": "cr:value", |
| "@type": "@json" |
| }, |
| "recordSet": "cr:recordSet", |
| "field": "cr:field", |
| "includes": "cr:includes", |
| "source": "cr:source", |
| "fileSet": "cr:fileSet", |
| "extract": "cr:extract", |
| "column": "cr:column" |
| }, |
| "@type": "sc:Dataset", |
| "name": "Poisson_Multimode PDEFieldDataset", |
| "description": "PosteriorBench v2.0 Hugging Face Arrow training dataset for the Poisson_Multimode inverse problem. Each of the 50,000 records contains an integer id, raw float32 forcing field a=f with shape [128, 128], raw float32 solution field u=phi with shape [128, 128], and scalar generation parameters alpha and tau. Runtime loaders concatenate a and u into channel-first [2, 128, 128] tensors and apply normalization from metadata.json.", |
| "conformsTo": [ |
| "http://mlcommons.org/croissant/1.1", |
| "http://mlcommons.org/croissant/RAI/1.0" |
| ], |
| "url": "https://huggingface.co/datasets/anonymousmay/PosteriorBench", |
| "license": "https://creativecommons.org/licenses/by/4.0/", |
| "creator": { |
| "@type": "Organization", |
| "name": "PosteriorBench / DDIS contributors", |
| "url": "https://github.com/neuraloperator/FunDPS-Physics-Guidance" |
| }, |
| "publisher": { |
| "@type": "Organization", |
| "name": "PosteriorBench / DDIS contributors", |
| "url": "https://github.com/neuraloperator/FunDPS-Physics-Guidance" |
| }, |
| "dateCreated": "2026-05-02", |
| "datePublished": "2026-05-05", |
| "dateModified": "2026-05-07", |
| "version": "2.0", |
| "keywords": [ |
| "inverse problems", |
| "partial differential equation", |
| "Poisson equation", |
| "posterior sampling", |
| "diffusion models", |
| "scientific machine learning", |
| "uncertainty quantification" |
| ], |
| "distribution": [ |
| { |
| "@type": "cr:FileSet", |
| "@id": "arrow-shards", |
| "includes": "data-*.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00000-of-00014.arrow", |
| "contentUrl": "data-00000-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "1cc3079b780150e67f626dd949f9797d831199791c1bd95e26e6d43d0135187f" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00001-of-00014.arrow", |
| "contentUrl": "data-00001-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "ae0c70111d9af7ab80374dd6dc233d563dc8d37098ac164e53a1bd9bcb3d4967" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00002-of-00014.arrow", |
| "contentUrl": "data-00002-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "20fa61d14065e2e5945ab420c5d42f30533c7e6148b76a3b004fda4d422e576e" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00003-of-00014.arrow", |
| "contentUrl": "data-00003-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "53693fa2b4e6dcb93c0dc49a167563b27c6452d35d970d1d8628132dd41f77c4" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00004-of-00014.arrow", |
| "contentUrl": "data-00004-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "bbf2e68b3f836076e91c14102d7a612efed653606ec6162081f026859a654cf0" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00005-of-00014.arrow", |
| "contentUrl": "data-00005-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "89eb2cbad51c5ab7c539136d7427103fe911d78fc5562232f60bb5563b42fbad" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00006-of-00014.arrow", |
| "contentUrl": "data-00006-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "26bade951e159f01b6973484a3a35476d627a281deea693cbc144ea323209997" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00007-of-00014.arrow", |
| "contentUrl": "data-00007-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "5da19aa088cb3bd5dc1f1d0b4413b59edf42c8f18785b7d0c269c97ff558e8a0" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00008-of-00014.arrow", |
| "contentUrl": "data-00008-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "61d9d69130e9a9395bded06b4f525fea9ff55da1d719839108073a0da3b761b6" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00009-of-00014.arrow", |
| "contentUrl": "data-00009-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "f68da5a1caea0ea611d6ed7e5a10def39c6ea7512c87cc708da324dc3e400339" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00010-of-00014.arrow", |
| "contentUrl": "data-00010-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "37393d44a83cef43ce80ced2c095018228100260f29cf13cfe6140e0a5dd69ed" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00011-of-00014.arrow", |
| "contentUrl": "data-00011-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "07fb3029387f50b809e54d27da6f39c36e1e8276d71a62e008303deed804cf8b" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00012-of-00014.arrow", |
| "contentUrl": "data-00012-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "7b77c240136bb363721ad19c2ed22cb27af4a10166a5d56d091aae45ed7e1e0f" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "data-00013-of-00014.arrow", |
| "contentUrl": "data-00013-of-00014.arrow", |
| "encodingFormat": "application/vnd.apache.arrow.stream", |
| "sha256": "b45a97c9012cf40c3186f851a537779fca165a54dfdf500c76e977f4c8332bc7" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "dataset_info.json", |
| "contentUrl": "dataset_info.json", |
| "encodingFormat": "application/json", |
| "sha256": "7add834c2c3129de2be7842105a023e737dfa16c318d3b5b4ea3243bb2ddf435" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "metadata.json", |
| "contentUrl": "metadata.json", |
| "encodingFormat": "application/json", |
| "sha256": "dd7bfc10fa9616b382c2545d22b5c031ee38c9e2ecca9de76d4c9d04275081dc" |
| }, |
| { |
| "@type": "cr:FileObject", |
| "@id": "state.json", |
| "contentUrl": "state.json", |
| "encodingFormat": "application/json", |
| "sha256": "53771678512f2fa5cfd156de42fb5c3d985b8792c960fef80d439ca9bcdbaa86" |
| } |
| ], |
| "rai:dataLimitations": "The dataset is synthetic Poisson PDE data at fixed 128x128 resolution and fixed alpha/tau parameter ranges. It should not be interpreted as physical measurements or as representative of all elliptic PDE inverse problems.", |
| "rai:dataBiases": "The generation metadata records alpha in [1.5, 3.0] and tau in [2.0, 4.0]; behavior outside those ranges is not represented.", |
| "rai:personalSensitiveInformation": "No personal or sensitive information is present.", |
| "rai:dataUseCases": [ |
| "Training diffusion priors and neural operators for Poisson PDE inverse-problem research.", |
| "Benchmarking posterior sampling, reconstruction, and scientific machine-learning methods on synthetic PDE fields.", |
| "Not recommended for claims about real-world physical systems without validation against domain data." |
| ], |
| "rai:dataSocialImpact": "Positive impact is expected primarily through reproducible scientific machine-learning benchmarks for inverse problems. No significant negative impacts are anticipated.", |
| "rai:hasSyntheticData": true, |
| "prov:wasDerivedFrom": [ |
| { |
| "@id": "urn:posteriorbench:source:poisson_multimode_params_h5", |
| "@type": "prov:Entity", |
| "name": "poisson_random_params.h5", |
| "description": "Raw conversion source containing f, phi, alpha, and tau arrays." |
| } |
| ], |
| "prov:wasGeneratedBy": [ |
| { |
| "@id": "urn:posteriorbench:activity:poisson_multimode_data_collection", |
| "@type": "prov:Activity", |
| "name": "Synthetic data collection", |
| "description": "The source fields were produced by a synthetic numerical Poisson PDE generator." |
| }, |
| { |
| "@id": "urn:posteriorbench:activity:poisson_multimode_hf_export", |
| "@type": "prov:Activity", |
| "name": "PosteriorBench v2.0 HF export", |
| "description": "The conversion stores raw physical fields under canonical names a and u, stores alpha/tau in params, records normalization statistics in metadata.json, and saves 50,000 rows as a Hugging Face Dataset backed by Apache Arrow shards." |
| }, |
| { |
| "@id": "urn:posteriorbench:activity:poisson_multimode_annotation", |
| "@type": "prov:Activity", |
| "name": "Annotation and labeling", |
| "description": "No human annotation was performed; semantic labels come from the numerical source field names." |
| } |
| ], |
| "recordSet": [ |
| { |
| "@type": "cr:RecordSet", |
| "@id": "poisson_multimode_train", |
| "field": [ |
| { |
| "@type": "cr:Field", |
| "@id": "poisson_multimode_train/split", |
| "dataType": "sc:Text", |
| "value": "train" |
| }, |
| { |
| "@type": "cr:Field", |
| "@id": "poisson_multimode_train/id", |
| "dataType": "sc:Integer", |
| "source": { |
| "fileSet": { |
| "@id": "arrow-shards" |
| }, |
| "extract": { |
| "column": "id" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "@id": "poisson_multimode_train/a", |
| "dataType": "cr:Float32", |
| "description": "Raw Poisson forcing field f with shape [128, 128].", |
| "source": { |
| "fileSet": { |
| "@id": "arrow-shards" |
| }, |
| "extract": { |
| "column": "a" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "@id": "poisson_multimode_train/u", |
| "dataType": "cr:Float32", |
| "description": "Raw Poisson solution field phi with shape [128, 128].", |
| "source": { |
| "fileSet": { |
| "@id": "arrow-shards" |
| }, |
| "extract": { |
| "column": "u" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "@id": "poisson_multimode_train/params/alpha", |
| "dataType": "cr:Float64", |
| "description": "GRF smoothness parameter alpha.", |
| "source": { |
| "fileSet": { |
| "@id": "arrow-shards" |
| }, |
| "extract": { |
| "column": "params.alpha" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "@id": "poisson_multimode_train/params/tau", |
| "dataType": "cr:Float64", |
| "description": "GRF scale parameter tau.", |
| "source": { |
| "fileSet": { |
| "@id": "arrow-shards" |
| }, |
| "extract": { |
| "column": "params.tau" |
| } |
| } |
| }, |
| { |
| "@type": "cr:Field", |
| "@id": "poisson_multimode_train/runtime_tensor_order", |
| "dataType": "sc:Text", |
| "value": [ |
| "a=f", |
| "u=phi" |
| ] |
| }, |
| { |
| "@type": "cr:Field", |
| "@id": "poisson_multimode_train/runtime_tensor_shape", |
| "dataType": "sc:Text", |
| "value": "PDEFieldDataset returns normalized concat(a, u) with shape [2, 128, 128]." |
| }, |
| { |
| "@type": "cr:Field", |
| "@id": "poisson_multimode_train/normalization", |
| "dataType": "sc:Text", |
| "value": "(x - mean) * (0.5 / std), with stats stored in metadata.json." |
| } |
| ] |
| } |
| ] |
| } |
|
|
