Publish ThousandWorlds v2.0.0

README.md

@@ -33,10 +33,9 @@ configs:
 
 <img src="imgs/MASCOT.png" align="right" width="220" style="margin-top: -1.25rem;" alt="ThousandWorlds mascot">
 
-ThousandWorlds is a benchmark for emulating exoplanet climates: **1760
-simulations** across **5 GCMs**, **8 planet parameters**, and atmospheric
+ThousandWorlds is a benchmark for emulating exoplanet climates: **1689 simulations** across **5 GCMs**, **8 planet parameters**, and atmospheric
 variables on a 32 x 64 x 10 latitude-longitude-pressure grid. It includes three
-nested benchmark subsets, two evaluation protocols, and eight released baseline
+nested benchmark subsets, two evaluation protocols, and ten released baseline
 methods.
 
 [![Code](https://img.shields.io/badge/code-GitHub-181717.svg?logo=github)](https://github.com/edstevenson/ThousandWorlds)
@@ -47,9 +46,6 @@ are time-averaged climate fields on a 32 x 64 latitude-longitude grid:
 three-dimensional variables are stored as pressure-level channels, and
 two-dimensional variables are stored as single-level fields.
 
-Want to look around before downloading anything? Try the [ThousandWorlds Explorer](https://thousandworldsexplorer.com)!
-Built by [Hamza Ali Shahjahan](https://github.com/hamza-ali-shahjahan)!
-
 ![ThousandWorlds dataset schematic](imgs/OVERVIEW.png)
 
 ## Quickstart
@@ -96,17 +92,17 @@ realism:
 
 | Subset | Simulations | Fields | Description |
 | --- | ---: | ---: | --- |
-| `single-complete` | 256 | 48 | Smaller subset; simulations from a single GCM, complete observations only. |
-| `multi-complete` | 1659 | 48 | All 5 GCMs, still with no missing fields. |
-| `multi-partial` | 1760 | 53 | Full dataset; all 5 GCMs, with missing fields represented as NaNs. |
+| `single-complete` | 277 | 48 | Smaller subset; simulations from a single GCM, complete observations only. |
+| `multi-complete` | 1595 | 48 | All 5 GCMs, still with no missing fields. |
+| `multi-partial` | 1689 | 53 | Full dataset; all 5 GCMs, with missing fields represented as NaNs. |
 
 The subset split files contain:
 
 | File | `single-complete` | `multi-complete` | `multi-partial` |
 | --- | ---: | ---: | ---: |
-| `train.csv` | 206 | 1538 | 1626 |
-| `test.csv` | 50 | 90 | 100 |
-| `test_shared_planets_only.csv` | - | 58 | 60 |
+| `train.csv` | 203 | 1443 | 1527 |
+| `test.csv` | 74 | 121 | 128 |
+| `test_shared_planets_only.csv` | - | 62 | 64 |
 | `held_out_aux.csv` | - | 31 | 34 |
 
 `held_out_aux.csv` is excluded from train and test to prevent train-test leakage (it contains simulations from auxiliary GCMs that correspond to identical planets present in the test set).
@@ -121,9 +117,9 @@ includes `is_target_gcm`, `in_target_physical_domain`, `planet_id`, and
 
 | Parameter | Range |
 | --- | --- |
-| Radius (Earth radii) | [0.7, 1.4] |
-| Surface gravity (m s^-2) | [6.0, 16.0] |
-| Rotation period (days) | [0.1, 1000.0] |
+| Radius (Earth radii) | [0.7, 1.75] |
+| Surface gravity (m s^-2) | [6.0, 17.0] |
+| Rotation period (days) | [1.0, 200.0] |
 | Surface pressure (bar) | [0.5, 5] |
 | CO2 volume fraction (%) | [0, 100] |
 | CH4 volume fraction (%) | [0, 5] |
@@ -153,10 +149,9 @@ The gridded field archives are:
 
 | File | Shape | Contents |
 | --- | --- | --- |
-| `dataset/fields/all-obs.npz` | `(1760, 53, 32, 64)` | Field archive covering all 5 GCMs with structured whole-field missingness. |
-| `dataset/fields/complete-obs-only.npz` | `(1659, 48, 32, 64)` | Complete-observation field archive. |
+| `dataset/fields/all-obs.npz` | `(1689, 53, 32, 64)` | Field archive covering all 5 GCMs with structured whole-field missingness. |
+| `dataset/fields/complete-obs-only.npz` | `(1595, 48, 32, 64)` | Complete-observation field archive. |
 
-**Spectral Coefficients:**
 The spectral coefficient archives mirror those field archives with T21
 spherical harmonic coefficients: `dataset/coefficients/*.npz` stores
 `coefficients` with 484 coefficients per field and a `field_mask` for missing
@@ -167,14 +162,13 @@ as false entries in the spectral `field_mask`.
 
 The package includes loaders and metrics for two benchmark protocols:
 
-- **Standard**: the main test protocol, ideal for ML model comparison.
+- **Standard**: the main test protocol.
 - **Shared-planets**: evaluate on planets shared across target and auxiliary
   GCMs; used to assess performance relative to inter-GCM error, i.e. how close
   a model gets to the epistemic uncertainty floor of the problem.
 
-Released baselines include train mean, kNN, PCA ridge, PCA-MLP, Coord-MLP,
-Coord-DeepONet, PPCA-ICM, and GPLFR. Baseline artifacts include predictions,
-resolved configs, and metrics JSON files.
+Headline baselines are `train_mean`, `knn`, `pca_gbt`, `coord_mlp`, `coord_deeponet`, `pca_mlp`, `conv_decoder`, `sfno`, `ppca_icm`, and `gplfr`. PCA-Ridge (`pca_ridge`) is the nonlinearity ablation. The baseline archives include predictions,
+the config each run used, and metrics JSON files.
 
 ## Links
 
@@ -196,3 +190,10 @@ If you use ThousandWorlds, please cite the paper:
   doi = {10.48550/arXiv.2606.18338}
 }
 ```
+
+## Versions
+
+v2.0.0 replaces v1.0.0. It is a small update to the dataset to fix a couple
+bugs in some simulations and add a few more; the full list is in the
+[changelog](https://github.com/edstevenson/ThousandWorlds/blob/main/CHANGELOG.md).
+v1.0.0 remains available for reproducing earlier work.

archives/dataset.tar.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:356c6cc14f6d23f6ffaef2155bfb668f6e365e07bb5b2f83736afa5343dae8b3
-size 869493997
+oid sha256:cddf50485fc32136f7742faf37b15c763a1340c21ff00b5689257064e77c252f
+size 837602811

archives/dataset.tar.gz.sha256

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:06210b7b77a6e2daba0a142b30760e3bf9007863289cc9831526e3ecd7254cd8
+oid sha256:e3eb406b9462f1d780d2962404368eda508b955b936766ccd827abd04972880a
 size 81

archives/results-baselines-multi-complete-deterministic.tar.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:204c492fd2af1b921946b1a9de07e39e8581c4f18cb01490b123bebf74cbc789
-size 123572271
+oid sha256:afa264f767531eb1059fd2bc3f84886e9c57f6f2e4b9cb02cc006554043fd730
+size 325855933

archives/results-baselines-multi-complete-deterministic.tar.gz.sha256

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:36f805332bf1c39a2e40a591a64155aa3f3e5b4218cf1f8ec2e7f98840373945
+oid sha256:d56b5d89b7a923598945b73ca0f4a12dfcfe37fd714bfbd4a8aa7326b84f1ec3
 size 120

archives/results-baselines-multi-complete-gplfr.tar.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:099018b9bb4c98fbc479889bd3a977fa249e0602fa8d7cd57f7a3ac78d42d356
-size 1150662113
+oid sha256:11fc79469e6d6213778f4a412739ea40147365cd7e588edd640b94b63c2100c0
+size 2353505563

archives/results-baselines-multi-complete-gplfr.tar.gz.sha256

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:68422681c6333cdd409e405d0a124f30e9c9c5dec0dee82067b79930aae84042
+oid sha256:59d31fd0e2cd138f13717466f048fab4536ee27d7cc0686c9a602e458f4b3698
 size 112

archives/results-baselines-multi-complete-ppca_icm.tar.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c3578059971aae6d3323b52dab27ad6f8133dcff7fefb8d79074a737200c21e3
-size 1158037580
+oid sha256:86e09a430859f0d59efc21146bb4f384051a651505052cada1f598709b2dd290
+size 2370310583

archives/results-baselines-multi-complete-ppca_icm.tar.gz.sha256

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a3673a921a5fb1d65018aeb064011be6d9b9b2e15c4a50666059d58025de8cdd
+oid sha256:0c12be6dc2d7dd0ff4b8d25252160e2ef591c37a2f75055a5f7e46d8e2a3e0d3
 size 115

archives/results-baselines-multi-partial-deterministic.tar.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:59f9979ed6c31fba1f43163485ba821e8cc486e1ca1893148528568b3a00b5ab
-size 151443554
+oid sha256:8b0da903b525f809f1902cd72bbdb6e7c944608d7e90565dc0e216f781fecc35
+size 376030722

archives/results-baselines-multi-partial-deterministic.tar.gz.sha256

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fcd36ca6e08477436f75b8c0cc66c8181db06de1c5d44013f7f81582996dc556
+oid sha256:456795efba6def79b2d898c8f35e020e359d018c98fbf5746f438a74e21c4f46
 size 119

archives/results-baselines-multi-partial-gplfr.tar.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:add9bcd8ba3761200ac89d013e259ebafeeb221b71043d062609ab3708ae2613
-size 1406475258
+oid sha256:cf35137a6b42c69e9d3c30c150d6a30f032cdd0fb8fe448f9065e2f78a66bee5
+size 2719870792

archives/results-baselines-multi-partial-gplfr.tar.gz.sha256

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7490b81090458451cb48f97f5ceec7de8d5957f0d9b561e1e892e3041ace338b
+oid sha256:b36d6737d6d4e8a56c841bb4d10f21cc3c034aaf1ac0d2f4737b14bcf46c3bed
 size 111

archives/results-baselines-multi-partial-ppca_icm.tar.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5d65fabd3662ce5da4aa0f3d1480285e633e6e1819e8c1a3e0f128b0017ae30a
-size 1411331865
+oid sha256:da7c817c17e02c5bb147222ffd0ea1734828ae586f38da48a3aa26cdfaaa8393
+size 2733193868

archives/results-baselines-multi-partial-ppca_icm.tar.gz.sha256

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a6eb12b2e692d080e20bba7c1c2ca7acc9c3199080feff2a0a454d7c8360334c
+oid sha256:4c2bf05a9b4bbd232321bf93f234384eb70886775ae94147eab95d9f5cf46be7
 size 114

archives/results-baselines-single-complete-deterministic.tar.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9725c008210b7c6bd5094a7707da0c305629fc36e9d533876ea85edde492ff9d
-size 69213653
+oid sha256:bafa76b3d197b089deea36970f6410a4803380b50a4c78756fa0e8dbf4cf1a37
+size 130968467

archives/results-baselines-single-complete-deterministic.tar.gz.sha256

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fad4c618c53159163a680f9ade91d47aed2387ad637d7d6f6726ab0d72580cc5
+oid sha256:488744837f34bb4bba825572355b814b90759a376a05a098f6001fee56a231b0
 size 121

archives/results-baselines-single-complete-gplfr.tar.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cf5a54f053836e0cd6758fa24c64d9d51ee1130c39831522dea12e8a59529a15
-size 646073153
+oid sha256:0dc0a22b092687fa25567a45a9749afcdcd5954844c4be2eba70a31ca138c571
+size 950625790

archives/results-baselines-single-complete-gplfr.tar.gz.sha256

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:02a43e8a5f6bc4a5f37a4ce8d90a5f913e53760062cd0cff359227fddbeb747d
+oid sha256:0cc7f31a62eddc44a3a3bf20ef558eeb9c580a54e2b6facff0895f97590dd3de
 size 113

archives/results-baselines-single-complete-ppca_icm.tar.gz

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:380331557807129ce722cc0884ec2518b7b334733fefc88e5e5797a1361860b5
-size 643746831
+oid sha256:2d7f2c4e56d9770a9ff027bc60895c196bb40d8444545a4cf19406e583bf891a
+size 944456994

archives/results-baselines-single-complete-ppca_icm.tar.gz.sha256

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c3e0c326582aec3680b353ec617dfa6f901b06057f647d5fe5a8a5d3509f5bbe
+oid sha256:70e0a8d92f4de9d8b167510535b400091431247626ae5168f9af002911d2ba72
 size 116

croissant.json

@@ -8,7 +8,6 @@
     "conformsTo": "dct:conformsTo",
     "containedIn": "cr:containedIn",
     "cr": "http://mlcommons.org/croissant/",
-    "rai": "http://mlcommons.org/croissant/RAI/",
     "data": {
       "@id": "cr:data",
       "@type": "@json"
@@ -28,8 +27,8 @@
     },
     "extract": "cr:extract",
     "field": "cr:field",
-    "fileProperty": "cr:fileProperty",
     "fileObject": "cr:fileObject",
+    "fileProperty": "cr:fileProperty",
     "fileSet": "cr:fileSet",
     "format": "cr:format",
     "includes": "cr:includes",
@@ -43,6 +42,7 @@
     },
     "parentField": "cr:parentField",
     "path": "cr:path",
+    "rai": "http://mlcommons.org/croissant/RAI/",
     "recordSet": "cr:recordSet",
     "references": "cr:references",
     "regex": "cr:regex",
@@ -55,250 +55,248 @@
     "subField": "cr:subField",
     "transform": "cr:transform"
   },
-  "@type": "sc:Dataset",
   "@id": "https://doi.org/10.57967/hf/8695",
-  "name": "ThousandWorlds",
-  "description": "Benchmark dataset and baseline result artifacts for exoplanet climate emulation. The release includes simulation input metadata, canonical train/test splits, gridded target fields, spectral coefficients, normalization statistics, baseline predictions, metrics, and per-variable score summaries.",
-  "url": "https://doi.org/10.57967/hf/8695",
-  "identifier": "https://doi.org/10.57967/hf/8695",
-  "license": "https://creativecommons.org/licenses/by/4.0/",
+  "@type": "sc:Dataset",
+  "auxiliaryMethods": [
+    "pca_ridge"
+  ],
+  "baselineMethods": [
+    "train_mean",
+    "knn",
+    "pca_gbt",
+    "coord_mlp",
+    "coord_deeponet",
+    "pca_mlp",
+    "conv_decoder",
+    "sfno",
+    "ppca_icm",
+    "gplfr"
+  ],
+  "citeAs": "Stevenson, E. T., Mak, M. T., Wolf, E., Sergeev, D. E., Hammond, T., Mayne, N. J., and Cranmer, M. ThousandWorlds: Benchmark dataset for exoplanet climate emulation. Hugging Face. https://doi.org/10.57967/hf/8695",
+  "conformsTo": "http://mlcommons.org/croissant/1.1",
   "creator": [
     {
       "@type": "sc:Person",
-      "name": "Edward T. Stevenson",
       "affiliation": {
         "@type": "sc:Organization",
         "name": "University of Cambridge"
       },
-      "email": "es833@cam.ac.uk"
+      "email": "es833@cam.ac.uk",
+      "name": "Edward T. Stevenson"
     },
     {
       "@type": "sc:Person",
-      "name": "Mei Ting Mak",
       "affiliation": {
         "@type": "sc:Organization",
         "name": "University of Oxford"
-      }
+      },
+      "name": "Mei Ting Mak"
     },
     {
       "@type": "sc:Person",
-      "name": "Eric Wolf",
       "affiliation": {
         "@type": "sc:Organization",
         "name": "University of Colorado Boulder"
-      }
+      },
+      "name": "Eric Wolf"
     },
     {
       "@type": "sc:Person",
-      "name": "Denis E. Sergeev",
       "affiliation": {
         "@type": "sc:Organization",
         "name": "University of Bristol"
-      }
+      },
+      "name": "Denis E. Sergeev"
     },
     {
       "@type": "sc:Person",
-      "name": "Tobi Hammond",
       "affiliation": {
         "@type": "sc:Organization",
         "name": "Purdue University"
-      }
+      },
+      "name": "Tobi Hammond"
     },
     {
       "@type": "sc:Person",
-      "name": "N. J. Mayne",
       "affiliation": {
         "@type": "sc:Organization",
         "name": "University of Exeter"
-      }
+      },
+      "name": "N. J. Mayne"
     },
     {
       "@type": "sc:Person",
-      "name": "Miles Cranmer",
       "affiliation": {
         "@type": "sc:Organization",
         "name": "University of Cambridge"
       },
-      "email": "mc2473@cam.ac.uk"
+      "email": "mc2473@cam.ac.uk",
+      "name": "Miles Cranmer"
     }
   ],
-  "keywords": [
-    "exoplanets",
-    "climate emulation",
-    "benchmark",
-    "surrogate modeling",
-    "probabilistic prediction"
-  ],
-  "datePublished": "2026-04-27",
-  "conformsTo": "http://mlcommons.org/croissant/1.1",
-  "rai:dataLimitations": "ThousandWorlds is a benchmark built from numerical exoplanet climate simulations, not observed planets. It is restricted to the released target physical domain, GCM choices, variables, spatial grid, temporal sampling, and benchmark split definitions, and should not be used as direct evidence for habitability, mission planning, or validated climate prediction outside those regimes.",
-  "rai:dataBiases": "The dataset reflects simulator, parameter-design, and preprocessing choices: ExoCAM and UM are target GCMs, auxiliary simulations are not uniformly represented across all benchmark subsets, and model behavior can depend on the released variable coverage and missing-field patterns.",
-  "rai:personalSensitiveInformation": "The dataset contains synthetic/numerical climate simulation outputs and associated physical parameters only. It contains no human-subject data and no personal or sensitive information.",
-  "rai:dataUseCases": "Validated use cases are benchmark development and evaluation for exoplanet climate emulation, surrogate modeling, missing-field handling, and probabilistic prediction over the released train/test protocols. The dataset is not validated for demographic fairness auditing, fine-tuning general-purpose AI systems, operational forecasting, or scientific claims beyond the documented simulation benchmark.",
-  "rai:dataSocialImpact": "Potential benefits include reproducible comparison of climate-emulation methods and broader access to expensive simulation-derived benchmark data. Main risks are overinterpreting benchmark scores or synthetic simulation outputs as observational truth; mitigations include explicit split definitions, public baseline artifacts, documented limitations, and open release metadata.",
-  "rai:hasSyntheticData": true,
+  "dataset_vintage": "tw-v2-aug20",
+  "datePublished": "2026-09-03",
+  "description": "Benchmark dataset and baseline result artifacts for exoplanet climate emulation. The release includes simulation input metadata, canonical train/test splits, gridded target fields, spectral coefficients, normalization statistics, baseline predictions, metrics, and per-variable score summaries.",
   "distribution": [
     {
-      "@type": "cr:FileObject",
       "@id": "dataset-archive",
-      "name": "dataset.tar.gz",
+      "@type": "cr:FileObject",
+      "contentUrl": "archives/dataset.tar.gz",
       "description": "Self-contained benchmark dataset archive. Extracts to dataset/ and includes metadata CSVs, split CSVs, gridded field NPZ archives, spectral coefficient NPZ archives, and normalization statistics.",
-      "contentUrl": "https://huggingface.co/datasets/es833/ThousandWorlds/resolve/v1.0.0/archives/dataset.tar.gz",
       "encodingFormat": "application/x-tar",
-      "sha256": "356c6cc14f6d23f6ffaef2155bfb668f6e365e07bb5b2f83736afa5343dae8b3"
+      "name": "dataset.tar.gz",
+      "sha256": "cddf50485fc32136f7742faf37b15c763a1340c21ff00b5689257064e77c252f"
     },
     {
-      "@type": "cr:FileObject",
       "@id": "results-baselines-multi-partial-deterministic",
-      "name": "results-baselines-multi-partial-deterministic.tar.gz",
+      "@type": "cr:FileObject",
+      "contentUrl": "archives/results-baselines-multi-partial-deterministic.tar.gz",
       "description": "Baseline result shard for the multi-partial-deterministic subset. Extracts to results/ and includes predictions, metrics JSON, paper tables, scores.csv, and results/README.md for this subset.",
-      "contentUrl": "https://huggingface.co/datasets/es833/ThousandWorlds/resolve/v1.0.0/archives/results-baselines-multi-partial-deterministic.tar.gz",
       "encodingFormat": "application/x-tar",
-      "sha256": "59f9979ed6c31fba1f43163485ba821e8cc486e1ca1893148528568b3a00b5ab"
+      "name": "results-baselines-multi-partial-deterministic.tar.gz",
+      "sha256": "8b0da903b525f809f1902cd72bbdb6e7c944608d7e90565dc0e216f781fecc35"
     },
     {
-      "@type": "cr:FileObject",
       "@id": "results-baselines-multi-complete-deterministic",
-      "name": "results-baselines-multi-complete-deterministic.tar.gz",
+      "@type": "cr:FileObject",
+      "contentUrl": "archives/results-baselines-multi-complete-deterministic.tar.gz",
       "description": "Baseline result shard for the multi-complete-deterministic subset. Extracts to results/ and includes predictions, metrics JSON, paper tables, scores.csv, and results/README.md for this subset.",
-      "contentUrl": "https://huggingface.co/datasets/es833/ThousandWorlds/resolve/v1.0.0/archives/results-baselines-multi-complete-deterministic.tar.gz",
       "encodingFormat": "application/x-tar",
-      "sha256": "204c492fd2af1b921946b1a9de07e39e8581c4f18cb01490b123bebf74cbc789"
+      "name": "results-baselines-multi-complete-deterministic.tar.gz",
+      "sha256": "afa264f767531eb1059fd2bc3f84886e9c57f6f2e4b9cb02cc006554043fd730"
     },
     {
-      "@type": "cr:FileObject",
       "@id": "results-baselines-single-complete-deterministic",
-      "name": "results-baselines-single-complete-deterministic.tar.gz",
+      "@type": "cr:FileObject",
+      "contentUrl": "archives/results-baselines-single-complete-deterministic.tar.gz",
       "description": "Baseline result shard for the single-complete-deterministic subset. Extracts to results/ and includes predictions, metrics JSON, paper tables, scores.csv, and results/README.md for this subset.",
-      "contentUrl": "https://huggingface.co/datasets/es833/ThousandWorlds/resolve/v1.0.0/archives/results-baselines-single-complete-deterministic.tar.gz",
       "encodingFormat": "application/x-tar",
-      "sha256": "9725c008210b7c6bd5094a7707da0c305629fc36e9d533876ea85edde492ff9d"
+      "name": "results-baselines-single-complete-deterministic.tar.gz",
+      "sha256": "bafa76b3d197b089deea36970f6410a4803380b50a4c78756fa0e8dbf4cf1a37"
     },
     {
-      "@type": "cr:FileObject",
       "@id": "results-baselines-multi-partial-gplfr",
-      "name": "results-baselines-multi-partial-gplfr.tar.gz",
+      "@type": "cr:FileObject",
+      "contentUrl": "archives/results-baselines-multi-partial-gplfr.tar.gz",
       "description": "Baseline result shard for the multi-partial-gplfr subset. Extracts to results/ and includes predictions, metrics JSON, paper tables, scores.csv, and results/README.md for this subset.",
-      "contentUrl": "https://huggingface.co/datasets/es833/ThousandWorlds/resolve/v1.0.0/archives/results-baselines-multi-partial-gplfr.tar.gz",
       "encodingFormat": "application/x-tar",
-      "sha256": "add9bcd8ba3761200ac89d013e259ebafeeb221b71043d062609ab3708ae2613"
+      "name": "results-baselines-multi-partial-gplfr.tar.gz",
+      "sha256": "cf35137a6b42c69e9d3c30c150d6a30f032cdd0fb8fe448f9065e2f78a66bee5"
     },
     {
-      "@type": "cr:FileObject",
       "@id": "results-baselines-multi-partial-ppca_icm",
-      "name": "results-baselines-multi-partial-ppca_icm.tar.gz",
+      "@type": "cr:FileObject",
+      "contentUrl": "archives/results-baselines-multi-partial-ppca_icm.tar.gz",
       "description": "Baseline result shard for the multi-partial-ppca_icm subset. Extracts to results/ and includes predictions, metrics JSON, paper tables, scores.csv, and results/README.md for this subset.",
-      "contentUrl": "https://huggingface.co/datasets/es833/ThousandWorlds/resolve/v1.0.0/archives/results-baselines-multi-partial-ppca_icm.tar.gz",
       "encodingFormat": "application/x-tar",
-      "sha256": "5d65fabd3662ce5da4aa0f3d1480285e633e6e1819e8c1a3e0f128b0017ae30a"
+      "name": "results-baselines-multi-partial-ppca_icm.tar.gz",
+      "sha256": "da7c817c17e02c5bb147222ffd0ea1734828ae586f38da48a3aa26cdfaaa8393"
     },
     {
-      "@type": "cr:FileObject",
       "@id": "results-baselines-multi-complete-gplfr",
-      "name": "results-baselines-multi-complete-gplfr.tar.gz",
+      "@type": "cr:FileObject",
+      "contentUrl": "archives/results-baselines-multi-complete-gplfr.tar.gz",
       "description": "Baseline result shard for the multi-complete-gplfr subset. Extracts to results/ and includes predictions, metrics JSON, paper tables, scores.csv, and results/README.md for this subset.",
-      "contentUrl": "https://huggingface.co/datasets/es833/ThousandWorlds/resolve/v1.0.0/archives/results-baselines-multi-complete-gplfr.tar.gz",
       "encodingFormat": "application/x-tar",
-      "sha256": "099018b9bb4c98fbc479889bd3a977fa249e0602fa8d7cd57f7a3ac78d42d356"
+      "name": "results-baselines-multi-complete-gplfr.tar.gz",
+      "sha256": "11fc79469e6d6213778f4a412739ea40147365cd7e588edd640b94b63c2100c0"
     },
     {
-      "@type": "cr:FileObject",
       "@id": "results-baselines-multi-complete-ppca_icm",
-      "name": "results-baselines-multi-complete-ppca_icm.tar.gz",
+      "@type": "cr:FileObject",
+      "contentUrl": "archives/results-baselines-multi-complete-ppca_icm.tar.gz",
       "description": "Baseline result shard for the multi-complete-ppca_icm subset. Extracts to results/ and includes predictions, metrics JSON, paper tables, scores.csv, and results/README.md for this subset.",
-      "contentUrl": "https://huggingface.co/datasets/es833/ThousandWorlds/resolve/v1.0.0/archives/results-baselines-multi-complete-ppca_icm.tar.gz",
       "encodingFormat": "application/x-tar",
-      "sha256": "c3578059971aae6d3323b52dab27ad6f8133dcff7fefb8d79074a737200c21e3"
+      "name": "results-baselines-multi-complete-ppca_icm.tar.gz",
+      "sha256": "86e09a430859f0d59efc21146bb4f384051a651505052cada1f598709b2dd290"
     },
     {
-      "@type": "cr:FileObject",
       "@id": "results-baselines-single-complete-gplfr",
-      "name": "results-baselines-single-complete-gplfr.tar.gz",
+      "@type": "cr:FileObject",
+      "contentUrl": "archives/results-baselines-single-complete-gplfr.tar.gz",
       "description": "Baseline result shard for the single-complete-gplfr subset. Extracts to results/ and includes predictions, metrics JSON, paper tables, scores.csv, and results/README.md for this subset.",
-      "contentUrl": "https://huggingface.co/datasets/es833/ThousandWorlds/resolve/v1.0.0/archives/results-baselines-single-complete-gplfr.tar.gz",
       "encodingFormat": "application/x-tar",
-      "sha256": "cf5a54f053836e0cd6758fa24c64d9d51ee1130c39831522dea12e8a59529a15"
+      "name": "results-baselines-single-complete-gplfr.tar.gz",
+      "sha256": "0dc0a22b092687fa25567a45a9749afcdcd5954844c4be2eba70a31ca138c571"
     },
     {
-      "@type": "cr:FileObject",
       "@id": "results-baselines-single-complete-ppca_icm",
-      "name": "results-baselines-single-complete-ppca_icm.tar.gz",
+      "@type": "cr:FileObject",
+      "contentUrl": "archives/results-baselines-single-complete-ppca_icm.tar.gz",
       "description": "Baseline result shard for the single-complete-ppca_icm subset. Extracts to results/ and includes predictions, metrics JSON, paper tables, scores.csv, and results/README.md for this subset.",
-      "contentUrl": "https://huggingface.co/datasets/es833/ThousandWorlds/resolve/v1.0.0/archives/results-baselines-single-complete-ppca_icm.tar.gz",
       "encodingFormat": "application/x-tar",
-      "sha256": "380331557807129ce722cc0884ec2518b7b334733fefc88e5e5797a1361860b5"
+      "name": "results-baselines-single-complete-ppca_icm.tar.gz",
+      "sha256": "2d7f2c4e56d9770a9ff027bc60895c196bb40d8444545a4cf19406e583bf891a"
     },
     {
-      "@type": "cr:FileObject",
       "@id": "inputs-csv",
-      "name": "dataset/inputs.csv",
-      "description": "Simulation input parameters and metadata inside dataset.tar.gz.",
-      "contentUrl": "dataset/inputs.csv",
+      "@type": "cr:FileObject",
       "containedIn": {
         "@id": "dataset-archive"
       },
-      "encodingFormat": "text/csv"
+      "contentUrl": "dataset/inputs.csv",
+      "description": "Simulation input parameters and metadata inside dataset.tar.gz.",
+      "encodingFormat": "text/csv",
+      "name": "dataset/inputs.csv"
     },
     {
-      "@type": "cr:FileSet",
       "@id": "split-csv-files",
-      "name": "dataset/subsets/*/*.csv",
-      "description": "Train, test, shared-planets test, and held-out auxiliary split membership CSVs inside dataset.tar.gz.",
+      "@type": "cr:FileSet",
       "containedIn": {
         "@id": "dataset-archive"
       },
+      "description": "Train, test, shared-planets test, and held-out auxiliary split membership CSVs inside dataset.tar.gz.",
       "encodingFormat": "text/csv",
-      "includes": "dataset/subsets/*/*.csv"
+      "includes": "dataset/subsets/*/*.csv",
+      "name": "dataset/subsets/*/*.csv"
     },
     {
-      "@type": "cr:FileSet",
       "@id": "field-archives",
-      "name": "dataset/fields/*.npz",
-      "description": "Gridded target field NPZ archives inside dataset.tar.gz.",
+      "@type": "cr:FileSet",
       "containedIn": {
         "@id": "dataset-archive"
       },
+      "description": "Gridded target field NPZ archives inside dataset.tar.gz.",
       "encodingFormat": "application/x-numpy",
-      "includes": "dataset/fields/*.npz"
+      "includes": "dataset/fields/*.npz",
+      "name": "dataset/fields/*.npz"
     },
     {
-      "@type": "cr:FileSet",
       "@id": "coefficient-archives",
-      "name": "dataset/coefficients/*.npz",
-      "description": "Spectral coefficient NPZ archives inside dataset.tar.gz.",
+      "@type": "cr:FileSet",
       "containedIn": {
         "@id": "dataset-archive"
       },
+      "description": "Spectral coefficient NPZ archives inside dataset.tar.gz.",
       "encodingFormat": "application/x-numpy",
-      "includes": "dataset/coefficients/*.npz"
+      "includes": "dataset/coefficients/*.npz",
+      "name": "dataset/coefficients/*.npz"
     },
     {
-      "@type": "cr:FileSet",
       "@id": "normalization-assets",
-      "name": "dataset/norm_stats/**/*",
-      "description": "Subset-specific normalization statistics and spectral transform metadata inside dataset.tar.gz.",
+      "@type": "cr:FileSet",
       "containedIn": {
         "@id": "dataset-archive"
       },
+      "description": "Subset-specific normalization statistics and spectral transform metadata inside dataset.tar.gz.",
       "encodingFormat": "application/octet-stream",
-      "includes": "dataset/norm_stats/**/*"
+      "includes": "dataset/norm_stats/**/*",
+      "name": "dataset/norm_stats/**/*"
     },
     {
-      "@type": "cr:FileObject",
       "@id": "scores-csv",
-      "name": "results/scores.csv",
-      "description": "Flat per-variable baseline score table included in the deterministic baseline archives and tracked in the source repository.",
-      "contentUrl": "results/scores.csv",
+      "@type": "cr:FileObject",
       "containedIn": {
         "@id": "results-baselines-single-complete-deterministic"
       },
-      "encodingFormat": "text/csv"
+      "contentUrl": "results/scores.csv",
+      "description": "Flat per-variable baseline score table included in the deterministic baseline archives and tracked in the source repository.",
+      "encodingFormat": "text/csv",
+      "name": "results/scores.csv"
     },
     {
-      "@type": "cr:FileSet",
       "@id": "baseline-predictions",
-      "name": "results/models/<subset>/<method>/predictions.npz",
-      "description": "Baseline prediction NPZ files inside the baseline result archives.",
+      "@type": "cr:FileSet",
       "containedIn": [
         {
           "@id": "results-baselines-multi-partial-deterministic"
@@ -328,14 +326,14 @@
           "@id": "results-baselines-single-complete-ppca_icm"
         }
       ],
+      "description": "Baseline prediction NPZ files inside the baseline result archives.",
       "encodingFormat": "application/x-numpy",
-      "includes": "results/models/<subset>/<method>/predictions.npz"
+      "includes": "results/models/<subset>/<method>/predictions.npz",
+      "name": "results/models/<subset>/<method>/predictions.npz"
     },
     {
-      "@type": "cr:FileSet",
       "@id": "baseline-metadata",
-      "name": "results/models/**/*.json",
-      "description": "Resolved baseline configs, metrics JSON files, and calibration metadata inside the baseline result archives.",
+      "@type": "cr:FileSet",
       "containedIn": [
         {
           "@id": "results-baselines-multi-partial-deterministic"
@@ -365,28 +363,40 @@
           "@id": "results-baselines-single-complete-ppca_icm"
         }
       ],
+      "description": "Resolved baseline configs, metrics JSON files, and calibration metadata inside the baseline result archives.",
       "encodingFormat": "application/json",
-      "includes": "results/models/**/*.json"
+      "includes": "results/models/**/*.json",
+      "name": "results/models/**/*.json"
     }
   ],
+  "identifier": "https://doi.org/10.57967/hf/8695",
+  "keywords": [
+    "exoplanets",
+    "climate emulation",
+    "benchmark",
+    "surrogate modeling",
+    "probabilistic prediction"
+  ],
+  "license": "https://creativecommons.org/licenses/by/4.0/",
+  "name": "ThousandWorlds",
+  "rai:dataBiases": "The dataset reflects simulator, parameter-design, and preprocessing choices: ExoCAM and UM are target GCMs, auxiliary simulations are not uniformly represented across all benchmark subsets, and model behavior can depend on the released variable coverage and missing-field patterns.",
+  "rai:dataLimitations": "ThousandWorlds is a benchmark built from numerical exoplanet climate simulations, not observed planets. It is restricted to the released target physical domain, GCM choices, variables, spatial grid, temporal sampling, and benchmark split definitions, and should not be used as direct evidence for habitability, mission planning, or validated climate prediction outside those regimes.",
+  "rai:dataSocialImpact": "Potential benefits include reproducible comparison of climate-emulation methods and broader access to expensive simulation-derived benchmark data. Main risks are overinterpreting benchmark scores or synthetic simulation outputs as observational truth; mitigations include explicit split definitions, public baseline artifacts, documented limitations, and open release metadata.",
+  "rai:dataUseCases": "Validated use cases are benchmark development and evaluation for exoplanet climate emulation, surrogate modeling, missing-field handling, and probabilistic prediction over the released train/test protocols. The dataset is not validated for demographic fairness auditing, fine-tuning general-purpose AI systems, operational forecasting, or scientific claims beyond the documented simulation benchmark.",
+  "rai:hasSyntheticData": true,
+  "rai:personalSensitiveInformation": "The dataset contains synthetic/numerical climate simulation outputs and associated physical parameters only. It contains no human-subject data and no personal or sensitive information.",
   "recordSet": [
     {
-      "@type": "cr:RecordSet",
       "@id": "inputs",
-      "name": "inputs",
+      "@type": "cr:RecordSet",
       "description": "One row per simulation, containing physical input parameters, GCM labels, split-relevant metadata, and provenance labels.",
-      "key": [
-        {
-          "@id": "inputs/simulation_id"
-        }
-      ],
       "field": [
         {
-          "@type": "cr:Field",
           "@id": "inputs/simulation_id",
-          "name": "simulation_id",
+          "@type": "cr:Field",
           "dataType": "sc:Integer",
           "description": "Integer simulation identifier used to index field archives, coefficient archives, split CSVs, and predictions.",
+          "name": "simulation_id",
           "source": {
             "extract": {
               "column": "simulation_id"
@@ -397,11 +407,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/radius",
-          "name": "radius",
+          "@type": "cr:Field",
           "dataType": "sc:Float",
           "description": "Planet radius in meters.",
+          "name": "radius",
           "source": {
             "extract": {
               "column": "radius"
@@ -412,11 +422,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/gravity",
-          "name": "gravity",
+          "@type": "cr:Field",
           "dataType": "sc:Float",
           "description": "Surface gravity in meters per second squared.",
+          "name": "gravity",
           "source": {
             "extract": {
               "column": "gravity"
@@ -427,11 +437,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/rotation_period",
-          "name": "rotation_period",
+          "@type": "cr:Field",
           "dataType": "sc:Float",
           "description": "Planet rotation period in Earth days.",
+          "name": "rotation_period",
           "source": {
             "extract": {
               "column": "rotation_period"
@@ -442,11 +452,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/surface_pressure",
-          "name": "surface_pressure",
+          "@type": "cr:Field",
           "dataType": "sc:Float",
           "description": "Surface pressure in pascals.",
+          "name": "surface_pressure",
           "source": {
             "extract": {
               "column": "surface_pressure"
@@ -457,11 +467,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/co2",
-          "name": "co2",
+          "@type": "cr:Field",
           "dataType": "sc:Float",
           "description": "Atmospheric CO2 setting used by the simulation.",
+          "name": "co2",
           "source": {
             "extract": {
               "column": "co2"
@@ -472,11 +482,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/ch4",
-          "name": "ch4",
+          "@type": "cr:Field",
           "dataType": "sc:Float",
           "description": "Atmospheric CH4 setting used by the simulation.",
+          "name": "ch4",
           "source": {
             "extract": {
               "column": "ch4"
@@ -487,11 +497,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/stellar_flux",
-          "name": "stellar_flux",
+          "@type": "cr:Field",
           "dataType": "sc:Float",
           "description": "Incident stellar flux in watts per square meter.",
+          "name": "stellar_flux",
           "source": {
             "extract": {
               "column": "stellar_flux"
@@ -502,11 +512,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/stellar_temperature",
-          "name": "stellar_temperature",
+          "@type": "cr:Field",
           "dataType": "sc:Float",
           "description": "Stellar effective temperature in kelvin.",
+          "name": "stellar_temperature",
           "source": {
             "extract": {
               "column": "stellar_temperature"
@@ -517,11 +527,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/gcm_label",
-          "name": "gcm_label",
+          "@type": "cr:Field",
           "dataType": "sc:Text",
           "description": "General circulation model label for the simulation.",
+          "name": "gcm_label",
           "source": {
             "extract": {
               "column": "gcm_label"
@@ -532,11 +542,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/is_target_gcm",
-          "name": "is_target_gcm",
+          "@type": "cr:Field",
           "dataType": "sc:Boolean",
           "description": "Whether the simulation is from a target GCM used in the main benchmark protocols.",
+          "name": "is_target_gcm",
           "source": {
             "extract": {
               "column": "is_target_gcm"
@@ -547,11 +557,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/in_target_physical_domain",
-          "name": "in_target_physical_domain",
+          "@type": "cr:Field",
           "dataType": "sc:Boolean",
           "description": "Whether the simulation lies inside the benchmark target physical domain.",
+          "name": "in_target_physical_domain",
           "source": {
             "extract": {
               "column": "in_target_physical_domain"
@@ -562,11 +572,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/planet_id",
-          "name": "planet_id",
+          "@type": "cr:Field",
           "dataType": "sc:Integer",
           "description": "Planet identifier used to group cross-GCM simulations of the same planet.",
+          "name": "planet_id",
           "source": {
             "extract": {
               "column": "planet_id"
@@ -577,11 +587,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "inputs/source",
-          "name": "source",
+          "@type": "cr:Field",
           "dataType": "sc:Text",
           "description": "Simulation provenance label.",
+          "name": "source",
           "source": {
             "extract": {
               "column": "source"
@@ -591,69 +601,55 @@
             }
           }
         }
-      ]
-    },
-    {
-      "@type": "cr:RecordSet",
-      "@id": "subsets",
-      "name": "subsets",
-      "description": "Canonical ThousandWorlds benchmark subsets.",
-      "dataType": "sc:Enumeration",
-      "key": {
-        "@id": "subsets/name"
-      },
-      "field": [
-        {
-          "@type": "cr:Field",
-          "@id": "subsets/name",
-          "name": "name",
-          "dataType": "sc:Text"
-        },
+      ],
+      "key": [
         {
-          "@type": "cr:Field",
-          "@id": "subsets/description",
-          "name": "description",
-          "dataType": "sc:Text"
+          "@id": "inputs/simulation_id"
         }
       ],
+      "name": "inputs"
+    },
+    {
+      "@id": "subsets",
+      "@type": "cr:RecordSet",
       "data": [
         {
-          "subsets/name": "multi-partial",
-          "subsets/description": "Broadest multi-GCM subset; allows structured whole-field missingness."
+          "subsets/description": "Broadest multi-GCM subset; allows structured whole-field missingness.",
+          "subsets/name": "multi-partial"
         },
         {
-          "subsets/name": "multi-complete",
-          "subsets/description": "Multi-GCM subset restricted to fields observed for every included simulation."
+          "subsets/description": "Multi-GCM subset restricted to fields observed for every included simulation.",
+          "subsets/name": "multi-complete"
         },
         {
-          "subsets/name": "single-complete",
-          "subsets/description": "Single-GCM subset restricted to complete observations."
+          "subsets/description": "Single-GCM subset restricted to complete observations.",
+          "subsets/name": "single-complete"
         }
-      ]
-    },
-    {
-      "@type": "cr:RecordSet",
-      "@id": "partitions",
-      "name": "partitions",
-      "description": "Canonical train/test partition names and their machine-readable ML partition semantics.",
-      "dataType": "cr:Split",
-      "key": {
-        "@id": "partitions/name"
-      },
+      ],
+      "dataType": "sc:Enumeration",
+      "description": "Canonical ThousandWorlds benchmark subsets.",
       "field": [
         {
+          "@id": "subsets/name",
           "@type": "cr:Field",
-          "@id": "partitions/name",
-          "name": "name",
-          "dataType": "sc:Text"
+          "dataType": "sc:Text",
+          "name": "name"
         },
         {
+          "@id": "subsets/description",
           "@type": "cr:Field",
-          "@id": "partitions/url",
-          "name": "url",
-          "dataType": "sc:Text"
+          "dataType": "sc:Text",
+          "name": "description"
         }
       ],
+      "key": {
+        "@id": "subsets/name"
+      },
+      "name": "subsets"
+    },
+    {
+      "@id": "partitions",
+      "@type": "cr:RecordSet",
       "data": [
         {
           "partitions/name": "train",
@@ -671,127 +667,129 @@
           "partitions/name": "held_out_aux",
           "partitions/url": "cr:TestSplit"
         }
-      ]
-    },
-    {
-      "@type": "cr:RecordSet",
-      "@id": "split_files",
-      "name": "split_files",
-      "description": "Canonical split membership CSV files. Each file contains one simulation_id column.",
-      "dataType": "sc:Enumeration",
-      "key": [
-        {
-          "@id": "split_files/subset"
-        },
-        {
-          "@id": "split_files/partition"
-        }
       ],
+      "dataType": "cr:Split",
+      "description": "Canonical train/test partition names and their machine-readable ML partition semantics.",
       "field": [
         {
+          "@id": "partitions/name",
           "@type": "cr:Field",
-          "@id": "split_files/subset",
-          "name": "subset",
-          "dataType": "sc:Text",
-          "description": "Benchmark subset."
-        },
-        {
-          "@type": "cr:Field",
-          "@id": "split_files/partition",
-          "name": "partition",
           "dataType": "sc:Text",
-          "description": "Split file stem."
+          "name": "name"
         },
         {
+          "@id": "partitions/url",
           "@type": "cr:Field",
-          "@id": "split_files/path",
-          "name": "path",
           "dataType": "sc:Text",
-          "description": "Path within dataset.tar.gz."
+          "name": "url"
         }
       ],
+      "key": {
+        "@id": "partitions/name"
+      },
+      "name": "partitions"
+    },
+    {
+      "@id": "split_files",
+      "@type": "cr:RecordSet",
       "data": [
         {
-          "split_files/subset": "multi-partial",
           "split_files/partition": "train",
-          "split_files/path": "dataset/subsets/multi-partial/train.csv"
+          "split_files/path": "dataset/subsets/multi-partial/train.csv",
+          "split_files/subset": "multi-partial"
         },
         {
-          "split_files/subset": "multi-partial",
           "split_files/partition": "test",
-          "split_files/path": "dataset/subsets/multi-partial/test.csv"
+          "split_files/path": "dataset/subsets/multi-partial/test.csv",
+          "split_files/subset": "multi-partial"
         },
         {
-          "split_files/subset": "multi-partial",
           "split_files/partition": "test_shared_planets_only",
-          "split_files/path": "dataset/subsets/multi-partial/test_shared_planets_only.csv"
+          "split_files/path": "dataset/subsets/multi-partial/test_shared_planets_only.csv",
+          "split_files/subset": "multi-partial"
         },
         {
-          "split_files/subset": "multi-partial",
           "split_files/partition": "held_out_aux",
-          "split_files/path": "dataset/subsets/multi-partial/held_out_aux.csv"
+          "split_files/path": "dataset/subsets/multi-partial/held_out_aux.csv",
+          "split_files/subset": "multi-partial"
         },
         {
-          "split_files/subset": "multi-complete",
           "split_files/partition": "train",
-          "split_files/path": "dataset/subsets/multi-complete/train.csv"
+          "split_files/path": "dataset/subsets/multi-complete/train.csv",
+          "split_files/subset": "multi-complete"
         },
         {
-          "split_files/subset": "multi-complete",
           "split_files/partition": "test",
-          "split_files/path": "dataset/subsets/multi-complete/test.csv"
+          "split_files/path": "dataset/subsets/multi-complete/test.csv",
+          "split_files/subset": "multi-complete"
         },
         {
-          "split_files/subset": "multi-complete",
           "split_files/partition": "test_shared_planets_only",
-          "split_files/path": "dataset/subsets/multi-complete/test_shared_planets_only.csv"
+          "split_files/path": "dataset/subsets/multi-complete/test_shared_planets_only.csv",
+          "split_files/subset": "multi-complete"
         },
         {
-          "split_files/subset": "multi-complete",
           "split_files/partition": "held_out_aux",
-          "split_files/path": "dataset/subsets/multi-complete/held_out_aux.csv"
+          "split_files/path": "dataset/subsets/multi-complete/held_out_aux.csv",
+          "split_files/subset": "multi-complete"
         },
         {
-          "split_files/subset": "single-complete",
           "split_files/partition": "train",
-          "split_files/path": "dataset/subsets/single-complete/train.csv"
+          "split_files/path": "dataset/subsets/single-complete/train.csv",
+          "split_files/subset": "single-complete"
         },
         {
-          "split_files/subset": "single-complete",
           "split_files/partition": "test",
-          "split_files/path": "dataset/subsets/single-complete/test.csv"
+          "split_files/path": "dataset/subsets/single-complete/test.csv",
+          "split_files/subset": "single-complete"
         }
-      ]
-    },
-    {
-      "@type": "cr:RecordSet",
-      "@id": "baseline_scores",
-      "name": "baseline_scores",
-      "description": "Flat per-variable baseline score table. One row corresponds to one subset, protocol, method, metric, and output variable.",
-      "key": [
+      ],
+      "dataType": "sc:Enumeration",
+      "description": "Canonical split membership CSV files. Each file contains one simulation_id column.",
+      "field": [
         {
-          "@id": "baseline_scores/subset"
+          "@id": "split_files/subset",
+          "@type": "cr:Field",
+          "dataType": "sc:Text",
+          "description": "Benchmark subset.",
+          "name": "subset"
         },
         {
-          "@id": "baseline_scores/protocol"
+          "@id": "split_files/partition",
+          "@type": "cr:Field",
+          "dataType": "sc:Text",
+          "description": "Split file stem.",
+          "name": "partition"
         },
         {
-          "@id": "baseline_scores/method"
-        },
+          "@id": "split_files/path",
+          "@type": "cr:Field",
+          "dataType": "sc:Text",
+          "description": "Path within dataset.tar.gz.",
+          "name": "path"
+        }
+      ],
+      "key": [
         {
-          "@id": "baseline_scores/metric"
+          "@id": "split_files/subset"
         },
         {
-          "@id": "baseline_scores/variable"
+          "@id": "split_files/partition"
         }
       ],
+      "name": "split_files"
+    },
+    {
+      "@id": "baseline_scores",
+      "@type": "cr:RecordSet",
+      "description": "Flat per-variable baseline score table. One row corresponds to one subset, protocol, method, metric, and output variable.",
       "field": [
         {
-          "@type": "cr:Field",
           "@id": "baseline_scores/subset",
-          "name": "subset",
+          "@type": "cr:Field",
           "dataType": "sc:Text",
           "description": "Benchmark subset.",
+          "name": "subset",
           "source": {
             "extract": {
               "column": "subset"
@@ -802,11 +800,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "baseline_scores/protocol",
-          "name": "protocol",
+          "@type": "cr:Field",
           "dataType": "sc:Text",
           "description": "Evaluation protocol.",
+          "name": "protocol",
           "source": {
             "extract": {
               "column": "protocol"
@@ -817,11 +815,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "baseline_scores/method",
-          "name": "method",
+          "@type": "cr:Field",
           "dataType": "sc:Text",
           "description": "Baseline method name.",
+          "name": "method",
           "source": {
             "extract": {
               "column": "method"
@@ -832,11 +830,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "baseline_scores/metric",
-          "name": "metric",
+          "@type": "cr:Field",
           "dataType": "sc:Text",
           "description": "Metric name, such as rmse, energy_score, relative_rmse, or relative_energy_score.",
+          "name": "metric",
           "source": {
             "extract": {
               "column": "metric"
@@ -847,11 +845,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "baseline_scores/variable",
-          "name": "variable",
+          "@type": "cr:Field",
           "dataType": "sc:Text",
           "description": "Output variable name after field-level metrics are aggregated to variables.",
+          "name": "variable",
           "source": {
             "extract": {
               "column": "variable"
@@ -862,11 +860,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "baseline_scores/value",
-          "name": "value",
+          "@type": "cr:Field",
           "dataType": "sc:Float",
           "description": "Metric value.",
+          "name": "value",
           "source": {
             "extract": {
               "column": "value"
@@ -877,11 +875,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "baseline_scores/metrics_path",
-          "name": "metrics_path",
+          "@type": "cr:Field",
           "dataType": "sc:Text",
           "description": "Path to the metrics JSON file within the extracted results tree.",
+          "name": "metrics_path",
           "source": {
             "extract": {
               "column": "metrics_path"
@@ -892,11 +890,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "baseline_scores/predictions_path",
-          "name": "predictions_path",
+          "@type": "cr:Field",
           "dataType": "sc:Text",
           "description": "Path to the prediction NPZ file within the extracted results tree.",
+          "name": "predictions_path",
           "source": {
             "extract": {
               "column": "predictions_path"
@@ -907,11 +905,11 @@
           }
         },
         {
-          "@type": "cr:Field",
           "@id": "baseline_scores/config_path",
-          "name": "config_path",
+          "@type": "cr:Field",
           "dataType": "sc:Text",
           "description": "Path to the resolved baseline config JSON file within the extracted results tree.",
+          "name": "config_path",
           "source": {
             "extract": {
               "column": "config_path"
@@ -921,9 +919,28 @@
             }
           }
         }
-      ]
+      ],
+      "key": [
+        {
+          "@id": "baseline_scores/subset"
+        },
+        {
+          "@id": "baseline_scores/protocol"
+        },
+        {
+          "@id": "baseline_scores/method"
+        },
+        {
+          "@id": "baseline_scores/metric"
+        },
+        {
+          "@id": "baseline_scores/variable"
+        }
+      ],
+      "name": "baseline_scores"
     }
   ],
-  "version": "1.0",
-  "citeAs": "Stevenson, E. T., Mak, M. T., Wolf, E., Sergeev, D. E., Hammond, T., Mayne, N. J., and Cranmer, M. ThousandWorlds: Benchmark dataset for exoplanet climate emulation. Hugging Face. https://doi.org/10.57967/hf/8695"
+  "release_status": "released",
+  "url": "https://doi.org/10.57967/hf/8695",
+  "version": "2.0.0"
 }

version.json

@@ -1,10 +1,92 @@
 {
-  "name": "ThousandWorlds",
-  "version": "1.0.0",
-  "hf_repo": "es833/ThousandWorlds",
-  "hf_revision": "v1.0.0",
   "archives": {
-    "dataset": "archives/dataset.tar.gz",
-    "baselines": "archives/results-baselines-*.tar.gz"
-  }
+    "dataset.tar.gz": {
+      "bytes": 837602811,
+      "path": "archives/dataset.tar.gz",
+      "sha256": "cddf50485fc32136f7742faf37b15c763a1340c21ff00b5689257064e77c252f",
+      "sha256_path": "archives/dataset.tar.gz.sha256"
+    },
+    "results-baselines-multi-complete-deterministic.tar.gz": {
+      "bytes": 325855933,
+      "path": "archives/results-baselines-multi-complete-deterministic.tar.gz",
+      "sha256": "afa264f767531eb1059fd2bc3f84886e9c57f6f2e4b9cb02cc006554043fd730",
+      "sha256_path": "archives/results-baselines-multi-complete-deterministic.tar.gz.sha256"
+    },
+    "results-baselines-multi-complete-gplfr.tar.gz": {
+      "bytes": 2353505563,
+      "path": "archives/results-baselines-multi-complete-gplfr.tar.gz",
+      "sha256": "11fc79469e6d6213778f4a412739ea40147365cd7e588edd640b94b63c2100c0",
+      "sha256_path": "archives/results-baselines-multi-complete-gplfr.tar.gz.sha256"
+    },
+    "results-baselines-multi-complete-ppca_icm.tar.gz": {
+      "bytes": 2370310583,
+      "path": "archives/results-baselines-multi-complete-ppca_icm.tar.gz",
+      "sha256": "86e09a430859f0d59efc21146bb4f384051a651505052cada1f598709b2dd290",
+      "sha256_path": "archives/results-baselines-multi-complete-ppca_icm.tar.gz.sha256"
+    },
+    "results-baselines-multi-partial-deterministic.tar.gz": {
+      "bytes": 376030722,
+      "path": "archives/results-baselines-multi-partial-deterministic.tar.gz",
+      "sha256": "8b0da903b525f809f1902cd72bbdb6e7c944608d7e90565dc0e216f781fecc35",
+      "sha256_path": "archives/results-baselines-multi-partial-deterministic.tar.gz.sha256"
+    },
+    "results-baselines-multi-partial-gplfr.tar.gz": {
+      "bytes": 2719870792,
+      "path": "archives/results-baselines-multi-partial-gplfr.tar.gz",
+      "sha256": "cf35137a6b42c69e9d3c30c150d6a30f032cdd0fb8fe448f9065e2f78a66bee5",
+      "sha256_path": "archives/results-baselines-multi-partial-gplfr.tar.gz.sha256"
+    },
+    "results-baselines-multi-partial-ppca_icm.tar.gz": {
+      "bytes": 2733193868,
+      "path": "archives/results-baselines-multi-partial-ppca_icm.tar.gz",
+      "sha256": "da7c817c17e02c5bb147222ffd0ea1734828ae586f38da48a3aa26cdfaaa8393",
+      "sha256_path": "archives/results-baselines-multi-partial-ppca_icm.tar.gz.sha256"
+    },
+    "results-baselines-single-complete-deterministic.tar.gz": {
+      "bytes": 130968467,
+      "path": "archives/results-baselines-single-complete-deterministic.tar.gz",
+      "sha256": "bafa76b3d197b089deea36970f6410a4803380b50a4c78756fa0e8dbf4cf1a37",
+      "sha256_path": "archives/results-baselines-single-complete-deterministic.tar.gz.sha256"
+    },
+    "results-baselines-single-complete-gplfr.tar.gz": {
+      "bytes": 950625790,
+      "path": "archives/results-baselines-single-complete-gplfr.tar.gz",
+      "sha256": "0dc0a22b092687fa25567a45a9749afcdcd5954844c4be2eba70a31ca138c571",
+      "sha256_path": "archives/results-baselines-single-complete-gplfr.tar.gz.sha256"
+    },
+    "results-baselines-single-complete-ppca_icm.tar.gz": {
+      "bytes": 944456994,
+      "path": "archives/results-baselines-single-complete-ppca_icm.tar.gz",
+      "sha256": "2d7f2c4e56d9770a9ff027bc60895c196bb40d8444545a4cf19406e583bf891a",
+      "sha256_path": "archives/results-baselines-single-complete-ppca_icm.tar.gz.sha256"
+    }
+  },
+  "auxiliary_methods": [
+    "pca_ridge"
+  ],
+  "dataset_vintage": "tw-v2-aug20",
+  "headline_methods": [
+    "train_mean",
+    "knn",
+    "pca_gbt",
+    "coord_mlp",
+    "coord_deeponet",
+    "pca_mlp",
+    "conv_decoder",
+    "sfno",
+    "ppca_icm",
+    "gplfr"
+  ],
+  "hf_repo": "es833/ThousandWorlds",
+  "hf_revision": "v2.0.0",
+  "name": "ThousandWorlds",
+  "release_status": "released",
+  "source_files": {
+    "README.md": "928bb3d9ad0c203474c69b37590a9dca722550ab4e0a094d823f76471f696676",
+    "croissant.json": "de1b7760b3992ca8b829de36aeaeae8c6513055634883fa8af7c7284ed8bd74e",
+    "imgs/MASCOT.png": "6060a16f7361686d29b7b7998544a4edbb41ee853ccc3eacf84964e7639c0ed8",
+    "imgs/OVERVIEW.png": "223d32ead45379fe1a422a26aa805b28857b803d85643e021e1b67dec6f797ab",
+    "inputs.csv": "31f355893974960f8d25a27d5a02bb1a70886387360a86b8880ea567eb11de95"
+  },
+  "version": "2.0.0"
 }