Add files using upload-large-folder tool

.gitattributes

@@ -1,60 +1,9 @@
-*.7z filter=lfs diff=lfs merge=lfs -text
-*.arrow filter=lfs diff=lfs merge=lfs -text
-*.avro filter=lfs diff=lfs merge=lfs -text
-*.bin filter=lfs diff=lfs merge=lfs -text
-*.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
-*.ftz filter=lfs diff=lfs merge=lfs -text
-*.gz filter=lfs diff=lfs merge=lfs -text
-*.h5 filter=lfs diff=lfs merge=lfs -text
-*.joblib filter=lfs diff=lfs merge=lfs -text
-*.lfs.* filter=lfs diff=lfs merge=lfs -text
-*.lz4 filter=lfs diff=lfs merge=lfs -text
-*.mds filter=lfs diff=lfs merge=lfs -text
-*.mlmodel filter=lfs diff=lfs merge=lfs -text
-*.model filter=lfs diff=lfs merge=lfs -text
-*.msgpack filter=lfs diff=lfs merge=lfs -text
-*.npy filter=lfs diff=lfs merge=lfs -text
-*.npz filter=lfs diff=lfs merge=lfs -text
-*.onnx filter=lfs diff=lfs merge=lfs -text
-*.ot filter=lfs diff=lfs merge=lfs -text
+# Hugging Face LFS — managed by `passage export`. Edit by regenerating with --force-gitattributes.
 *.parquet filter=lfs diff=lfs merge=lfs -text
-*.pb filter=lfs diff=lfs merge=lfs -text
-*.pickle filter=lfs diff=lfs merge=lfs -text
-*.pkl filter=lfs diff=lfs merge=lfs -text
+*.npy.zst filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
 *.pt filter=lfs diff=lfs merge=lfs -text
-*.pth filter=lfs diff=lfs merge=lfs -text
-*.rar filter=lfs diff=lfs merge=lfs -text
 *.safetensors filter=lfs diff=lfs merge=lfs -text
-saved_model/**/* filter=lfs diff=lfs merge=lfs -text
-*.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
-*.tflite filter=lfs diff=lfs merge=lfs -text
-*.tgz filter=lfs diff=lfs merge=lfs -text
-*.wasm filter=lfs diff=lfs merge=lfs -text
-*.xz filter=lfs diff=lfs merge=lfs -text
-*.zip filter=lfs diff=lfs merge=lfs -text
-*.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text
-# Audio files - uncompressed
-*.pcm filter=lfs diff=lfs merge=lfs -text
-*.sam filter=lfs diff=lfs merge=lfs -text
-*.raw filter=lfs diff=lfs merge=lfs -text
-# Audio files - compressed
-*.aac filter=lfs diff=lfs merge=lfs -text
-*.flac filter=lfs diff=lfs merge=lfs -text
-*.mp3 filter=lfs diff=lfs merge=lfs -text
-*.ogg filter=lfs diff=lfs merge=lfs -text
-*.wav filter=lfs diff=lfs merge=lfs -text
-# Image files - uncompressed
-*.bmp filter=lfs diff=lfs merge=lfs -text
-*.gif filter=lfs diff=lfs merge=lfs -text
-*.png filter=lfs diff=lfs merge=lfs -text
-*.tiff filter=lfs diff=lfs merge=lfs -text
-# Image files - compressed
-*.jpg filter=lfs diff=lfs merge=lfs -text
-*.jpeg filter=lfs diff=lfs merge=lfs -text
-*.webp filter=lfs diff=lfs merge=lfs -text
-# Video files - compressed
-*.mp4 filter=lfs diff=lfs merge=lfs -text
-*.webm filter=lfs diff=lfs merge=lfs -text

1024x1024_samples/calibration/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:058c329f3f1d90315e7b5bc5cb03e74fcc570b727e9d1a41b59ab0fbeec367db
+size 167014947

1024x1024_samples/test/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a680acfeacf9266cfc6b44cf8f0732060f294501ea2547959ad4a92b9d424f7d
+size 184248364

1024x1024_samples/train/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10f5746035c4afdf63200fe7b77adaf0215c1a7cf2db0def967fc2a1fbd23f5e
+size 174305821

1024x1024_samples/validation/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5dd766a1a625c3751337f3eecee11fa330aac937937b1dbedcca7962f058588f
+size 172274722

128x128_samples/calibration/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75d329f076d749b5b468c7a69d8e9725c7d576b29e92d837259c6b8afed0c0d6
+size 4387435

128x128_samples/test/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c111f3a7bb1207da77864c68fb3f6858f9588552cae82e35e8b54a2cc64c998a
+size 4882875

128x128_samples/train/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:adfd7756580df68705034160c45343b5678e379b9d38c88598b1a04ea58e00b3
+size 4647544

128x128_samples/validation/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fc6df21092cb8ed1a014d38ce66e0b00829e1948bbc450e02c88b80c2d3b24f
+size 4221782

2048x2048_samples/calibration/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d4fe1379246f949ddd65cd31525e55a305432452ce024719adda7baa3f7af43
+size 631609829

2048x2048_samples/test/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e0010b8378e3ba728ee0ba3566e7a6a2f12be909d6bcf3a5b59fcc1b9642e87c
+size 686239294

2048x2048_samples/train/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4754c996cdcb2f0335f393d7b7f0c5b9f0f770cf5f2f741d5fbb70a263de491d
+size 665460888

2048x2048_samples/validation/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dea08f019b4b91d6ae8cfe4b5020936faa88cb3a78c9c221a14c1da77fcae281
+size 588256216

256x256_samples/calibration/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8be604ceabe8f649339c0d24ff2d2bbd4a1099073f2d9c8fd7d1a23c846cec47
+size 14654757

256x256_samples/test/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9eb4d85bf1b42a83215362daf0b198500a0fdbb5ab380321260a9df47d16ff75
+size 16033785

256x256_samples/train/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:963b3141e75c82e75385beebd720d557946f2d287c94e231b0c852ab41ecb0a5
+size 14356660

256x256_samples/validation/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7467fa3c4ab3be11eea39b9dbafb81e905244187b56628cb78fdc637d348b638
+size 15008854

4096x4096_samples/calibration/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b350b35aee98f1a1777c1548d17a902cd56953047b4358262c5d95f0a85b2e7d
+size 956178608

4096x4096_samples/calibration/data-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:132ea543a62545022741d9cae12582b2b200c3856c24235b99c4989a78e11fab
+size 885212892

4096x4096_samples/calibration/data-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:36787c69135b91a47a22c3041265cca329f5cd958768d82a62b3bf1327f04664
+size 481018747

4096x4096_samples/test/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4916846005cbc380fdd1b7f7f28265b684cf3d6d10dc63207829caa921ad3019
+size 937015759

4096x4096_samples/test/data-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:45d9e2afdc733a048f3992acda270875de5ad25ddd92e3ad89857f67ff0f94eb
+size 846975711

4096x4096_samples/test/data-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f465a8e934d977ea5a391bcdf65c40b543dfa9b54c7a063ecb5e2ab84da3754a
+size 492585763

4096x4096_samples/train/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b066640435c2778f91a988097c58434fd6ddc09dca03790060ae71f01030a282
+size 899155512

4096x4096_samples/train/data-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d2ba55ac1c1f62dd04e2f71686b6088af692a0f20f7739b0de174013045d757b
+size 885102421

4096x4096_samples/train/data-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2fd9362a71444cf6b8c3a14b9b56a9d5d12705fc0885121a470329633e0e82ce
+size 498174779

4096x4096_samples/validation/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9ea624acfc9920923cda3fae02e41b5440e3e232a55f322ee15c6498d81870df
+size 918008866

4096x4096_samples/validation/data-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d0b475d03a21cf202b6348585c07e7242d4fe334fa76f86a36a4c1101a810c54
+size 844220976

4096x4096_samples/validation/data-00002.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e9bd7a6d8aca786c10e774123c06775ff9d3f79fbf0fe99180202edc3d21c8f2
+size 469486611

512x512_samples/calibration/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:76abe33ade8614046b9c36209f3ecacfe3ece843f1f80e3bda5765d4455e75c2
+size 48198259

512x512_samples/test/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c28637f17e74f319d5ec86047b6f0984d798c550994b61b344135ffc10a9579
+size 55770578

512x512_samples/train/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d33a31589913b5665098aeee494a541f87bd604a2d230c30a753d379255ea782
+size 48200926

512x512_samples/validation/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59d4ad045cc4c820bc6b808e20aaa4ef735b81e864b677b0f01d63bd3cb0a79a
+size 49476086

64x64_samples/calibration/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:180cc0a0546b083f3ee38ea98866956fb246259fcd896f360e620286081d1c3b
+size 1478003

64x64_samples/test/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b820a51627578357c25a932a697c60b00a2ab3d3c43c7cb3d9f95946e1c4e10
+size 1592634

64x64_samples/train/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75563398275d5d6d234db931bbf718c6f7fdda8ebba975aa0b1c774586ded51f
+size 1464420

64x64_samples/validation/data-00000.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a88a84c65d927eb00f536310503d2097294c98b119779c836ee0316e25f7edcb
+size 1442910

CHANGELOG.md

@@ -0,0 +1,17 @@
+# Changelog
+
+## Unreleased
+
+### Added
+
+- Deterministic per-sample generation seeding via `generation.seed`, with recorded `generation_seed` and `sample_seed` provenance in sample metadata.
+- Repository release metadata files: `CITATION.cff`, `DATASHEET.md`, `croissant.json`, and `Makefile`.
+- `make manifest` workflow for writing `outputs/export/manifest.sha256` from a prepared export tree.
+- `DATASET_LICENSE.md` to document the dataset-side JAXA AW3D30 obligations separately from the MIT code license.
+
+### Changed
+
+- Corrected public documentation to reflect the default 4,400,000-sample (≈4.4 M) benchmark size — 1,000,000 per resolution at 64–512, 250,000 at 1024, 100,000 at 2048, 50,000 at 4096 — and the current reproducibility contract.
+- Exposed deterministic-generation provenance in the export schema used for Hugging Face parquet preparation.
+- Restored `config/config.yaml` to the 4.4 M-sample paper benchmark and aligned preview-image defaults with the sample export settings.
+- Made `passage export` stage release metadata files into the export tree, with anonymous submission copies available via `--anonymous`.

CITATION.cff

@@ -0,0 +1,37 @@
+cff-version: 1.2.0
+title: PASSAGE
+message: If you use PASSAGE, please cite the software repository and accompanying dataset metadata in this file.
+type: software
+authors:
+  - family-names: Anonymous
+    given-names: Anonymous
+    email: anonymous@anonymous.invalid
+    affiliation: Anonymous Group
+  - family-names: Anonymous
+    given-names: Anonymous
+    affiliation: Anonymous Group
+  - family-names: Anonymous
+    given-names: Anonymous
+version: 0.1.0
+date-released: 2026-03-30
+license: MIT
+repository-code: https://github.com/anonymous-org/passage
+url: https://github.com/anonymous-org/passage
+keywords:
+  - pathfinding
+  - terrain-routing
+  - benchmark
+  - digital-elevation-model
+  - dataset
+preferred-citation:
+  type: article
+  title: PASSAGE: A Multi-Resolution Pathfinding Dataset Mixing Complex Obstacles and Elevation-Based Tiles Towards Certified Machine Learning Systems
+  authors:
+    - family-names: Anonymous
+      given-names: Anonymous
+    - family-names: Anonymous
+      given-names: Anonymous
+    - family-names: Anonymous
+      given-names: Anonymous
+  year: 2026
+  url: https://github.com/anonymous-org/passage

DATASET_LICENSE.md

@@ -0,0 +1,19 @@
+# PASSAGE Dataset Terms
+
+PASSAGE ships two distinct licensing surfaces:
+
+- The generator code in this repository is released under the MIT license in `LICENSE`.
+- The exported dataset artifacts inherit obligations from the upstream JAXA ALOS AW3D30 elevation product used to derive the terrain channel.
+
+For dataset artifacts distributed on Hugging Face and in release bundles:
+
+- Preserve attribution to JAXA ALOS World 3D - 30m (AW3D30) and any other provenance notices carried in the metadata.
+- Do not represent the derived elevation-backed artifacts as public domain.
+- Comply with the upstream AW3D30 usage terms for any redistribution or downstream reuse of the derived terrain data.
+- Treat PASSAGE as a research benchmark, not a safety-of-life product. The dataset and generated paths do not certify operational deployment.
+
+When in doubt:
+
+- Use `LICENSE` for code questions.
+- Use this file plus `DATASHEET.md` for dataset-distribution questions.
+- Preserve both files when mirroring the release tree.

DATASHEET.md

@@ -0,0 +1,191 @@
+# PASSAGE Datasheet
+
+This Datasheet follows the Gebru et al. [*Datasheets for Datasets*](https://arxiv.org/abs/1803.09010) template and is mirrored in the companion Croissant metadata file (`croissant.json`, conforming to MLCommons Croissant 1.1 and Croissant RAI 1.0). It accompanies the NeurIPS 2026 Evaluations & Datasets Track submission of the PASSAGE benchmark.
+
+## Motivation
+
+### For what purpose was the dataset created?
+
+PASSAGE was created to close a reproducibility gap in *constrained path planning on real terrain*: existing benchmarks either synthesize grids that strip geographic structure, rely on ad-hoc GIS workflows without machine-learning-ready splits, or evaluate embodied perception rather than cost-aware long-horizon routing. PASSAGE provides a real-terrain, multi-resolution benchmark with an enumerable Operational Design Domain (ODD), configurable terrain cost models, procedural obstacles, and exact A* ground-truth labels — packaged as ML-ready Parquet shards with a manifest-checked release tree.
+
+### Who created this dataset and on behalf of which entity?
+
+The dataset was created by Anonymous Author, Anonymous Author, and Anonymous Author on behalf of Anonymous Group research. See `CITATION.cff` for full attribution. Maintenance is owned by the Anonymous Group open-source research SIG; community contact via GitHub Issues, Discussions, or `oss@anonymous.invalid`.
+
+### Who funded the creation of the dataset?
+
+Internal research funding (Anonymous Group). No external grant is associated with this release.
+
+## Composition
+
+### What do the instances represent?
+
+Each instance is a single path-planning problem on a terrain crop: a three-channel input tensor (elevation, start/goal markers, obstacle mask) at a chosen resolution `N`, plus path annotations (ordered `(row, col)` waypoint lists under each configured cost model), a binary `N×N` path mask per cost model, and structured metadata (source tile, crop bounds, resolution, cost model, weight, obstacle configuration, split, solver backend, path cost, path length, solver wall-clock, seeds).
+
+### How many instances are there?
+
+The default release emits **4,400,000 samples** (≈4.4 M) across seven resolutions (64, 128, 256, 512, 1024, 2048, 4096), with per-resolution targets `1,000,000` for 64–512, `250,000` for 1024, `100,000` for 2048, and `50,000` for 4096 (see `config/config.yaml::num_samples`). Per-resolution, per-split counts (`90/2/4/4` train/calibration/validation/test) are recorded in `outputs/paper/results/dataset_statistics.csv` and reported in Table 2 of the paper.
+
+### Does the dataset contain all possible instances or is it a sample?
+
+PASSAGE is a deterministically seeded *sample* from an enumerable parameter space. Without obstacles, the ODD is finite and fully enumerable: every `(tile, crop, resolution, cost model, weight, start, goal)` maps deterministically to a unique sample. Additional samples can be generated at will with `make rebuild`.
+
+### What data does each instance consist of?
+
+- `elevation`: real-valued, normalized against globally calibrated min/max.
+- `markers`: binary start/goal channels.
+- `obstacles`: binary forbidden-cell mask (optional; the free-terrain variant is `obstacles=0`).
+- `path_waypoints_<cost_model>`: ordered `(row, col)` pairs.
+- `path_mask_<cost_model>`: binary `N×N`.
+- `metadata`: structured provenance (see Croissant `record_set.metadata`).
+
+### Is there a label or target associated with each instance?
+
+Yes — reference paths under each cost model. Paths are *computational annotations* produced by the A* solver (grid backend, Numba-JIT), not human annotations. There are no ambiguous or missing labels by construction.
+
+### Is any information missing from individual instances?
+
+No; every sample is fully populated by the generator. Upstream AW3D30 tiles may carry sensor artifacts (voids, striping); affected samples are flagged in metadata via the source-tile identifier and, where relevant, the elevation calibration bounds.
+
+### Are relationships between individual instances made explicit?
+
+Yes: samples inherit the `source_tile_id` and `split` fields from their source AW3D30 tile. The geographic hold-out split policy (`split_mask` at 1°/pixel) is versioned alongside the generator configuration.
+
+### Are there recommended data splits?
+
+Yes — the default split is 90% train / 2% calibration / 4% validation / 4% test, with a *geographic hold-out* for test (entire 1°×1° source tiles are reserved for test; no source tile appears in both train and test).
+
+### Are there any errors, sources of noise, or redundancies?
+
+The elevation product inherits JAXA ALOS AW3D30 sensor characteristics (DSM error bounds documented upstream). Path annotations are exact under the documented cost model — no stochastic error. Procedural obstacles are abstractions of real hazards and are not validated against real-world obstacle maps.
+
+### Is the dataset self-contained, or does it link to other resources?
+
+Fully self-contained at the export stage. Raw AW3D30 tiles are re-derivable via `make download` but not shipped in the release.
+
+### Does the dataset contain confidential, personally identifiable, sensitive, or offensive information?
+
+No. PASSAGE contains no personal data, no biometric data, and no human subjects. Offensive content is not possible in this domain.
+
+## Collection Process
+
+### How was the data acquired?
+
+Raw elevation tiles are downloaded from the public JAXA ALOS AW3D30 archive (5°×5° packaging) and extracted into 1°×1° sub-tiles at native 30 m/pixel resolution. A global calibration pass computes elevation min/max across the local tile cache. For each requested `(resolution, sample_idx, split)` triple, a deterministic `blake2b` seed drives tile selection, crop placement, start/goal placement, obstacle synthesis, and cost-weight draw. Reference paths are solved with the Numba-JIT grid-backend A* under each configured cost model (see `src/passage/pathfinding_utils.py`).
+
+### What mechanisms or procedures were used to collect the data?
+
+Code-driven pipeline end-to-end (see `Makefile` and `config/config.yaml`). No crowdsourcing, no human annotators, no subject interaction.
+
+### Was any preprocessing, cleaning, or labeling done?
+
+- Elevation normalised against globally calibrated min/max.
+- Markers placed subject to obstacle-exclusion and minimum separation.
+- Obstacle masks synthesised from log-uniform super-ellipse parameters, rejected if coverage exceeds 30% after up to 50 placement attempts.
+- Paths solved with exact A* under the configured cost model and the 8-connected grid (step lengths 30 m and 30√2 m).
+
+### Was the "raw" data saved in addition to the preprocessed data?
+
+Yes — the upstream AW3D30 archive and the calibrated tile cache are persisted locally under `/data/PASSAGE/download/` and `/data/PASSAGE/calibrate/` respectively. Only the processed export (Parquet shards) is distributed.
+
+### Is the software available for the preprocessing?
+
+Yes — the full generator is open-source in this repository (`src/passage/`), with CI coverage and an end-to-end rebuild via `make rebuild`.
+
+## Uses
+
+### Has the dataset been used for any tasks already?
+
+Yes — PASSAGE is the primary benchmark for the downstream `passage-vision` workspace, which reports dense-field segmentation and cost-to-go regression baselines at the 256×256 operating point, plus a 64→4096 resolution-frontier study.
+
+### What (other) tasks could the dataset be used for?
+
+- Benchmarking classical path-planning algorithms (Dijkstra, Theta*, D\*, JPS) on real terrain.
+- Training and evaluating graph-neural planners and RL planners in bridge comparisons.
+- Studying multi-resolution scaling, geographic distribution shift, and cost-function transfer in dense prediction.
+- Reproducible cost-model design studies with physically interpretable terms.
+
+### Is there anything about the composition of the dataset or the way it was collected that might impact future uses?
+
+- Samples are grid-based; they do not support full 3D motion planning or dynamic environments.
+- Obstacles are synthetic. Conclusions about real-world obstacle distributions cannot be drawn directly.
+- Geographic coverage is bounded by AW3D30 availability.
+
+### Are there tasks for which the dataset should not be used?
+
+- **Operational airborne flight planning or search-and-rescue dispatch**: PASSAGE is an advisory research benchmark. It does not by itself validate safety-critical deployment and does not substitute for EASA/FAA certification processes.
+- **Conclusions about weather, airspace, vegetation, urban structures, or regulations**: out of scope.
+- **Safety-of-life decisions**: predicted paths are not safety-of-life outputs without an in-the-loop deterministic verifier and fallback solver.
+
+## Distribution
+
+### Will the dataset be distributed to third parties?
+
+Yes — the dataset is distributed publicly via Hugging Face (`https://huggingface.co/datasets/Anonym-2045/passage`). For the NeurIPS 2026 anonymous submission, a reviewer-accessible anonymous mirror is provided and documented in the submission package.
+
+### How will it be distributed?
+
+Parquet shards, zstandard-compressed, one file per resolution × split. Accompanied by:
+
+- `croissant.json` (MLCommons 1.1 + RAI 1.0).
+- `outputs/export/manifest.sha256` checksum manifest.
+- README on Hugging Face mirroring the GitHub README.
+- `CITATION.cff`, `CHANGELOG.md`, `LICENSE`, this Datasheet.
+
+### When will it be distributed?
+
+Version 0.1.0 is released with the NeurIPS 2026 submission (May 2026). Future versions follow semantic versioning in `CHANGELOG.md`.
+
+### Will the dataset be distributed under a license? Will it have an associated DOI?
+
+- Code: MIT (see `LICENSE`).
+- Dataset: distributed under `DATASET_LICENSE.md`, which documents the inherited JAXA ALOS AW3D30 obligations for the derived artifacts.
+- A Zenodo DOI is planned for the v1.0 release (camera-ready gate) and will be recorded in `CITATION.cff`.
+
+### Have any third parties imposed IP-based or other restrictions?
+
+JAXA ALOS AW3D30 usage terms are inherited by the derived elevation channel. Users of PASSAGE must comply with the upstream JAXA terms, including attribution.
+
+### Do any export controls or other regulatory restrictions apply?
+
+No export controls apply; AW3D30 is a public global elevation product.
+
+## Maintenance
+
+### Who is supporting / hosting / maintaining the dataset?
+
+Anonymous Group open-source and critical AI research contributors. Contact: `oss@anonymous.invalid`; issues and discussions on the GitHub repository.
+
+### How can the owner be contacted?
+
+GitHub Issues and Discussions at <https://github.com/anonymous-org/passage>, or email `oss@anonymous.invalid`.
+
+### Is there an erratum?
+
+Errata will be tracked in `CHANGELOG.md`. Breaking schema changes bump the minor or major version and invalidate prior manifest checksums.
+
+### Will the dataset be updated?
+
+Yes. Updates are published via Hugging Face with a corresponding semantic-version tag in this repository and an entry in `CHANGELOG.md`.
+
+### If the dataset relates to people: are there applicable limits on the retention of the data associated with the instances?
+
+Not applicable — no human subjects.
+
+### Will older versions of the dataset continue to be supported/hosted/maintained?
+
+Prior releases remain accessible via their Hugging Face dataset-repo commit SHA and the git tag on GitHub. Deprecations are announced in `CHANGELOG.md`. The deprecation policy: one-minor-version notice, with `rai:dataReleaseMaintenancePlan` documenting active supported versions.
+
+### If others want to extend / augment / build on / contribute to the dataset, is there a mechanism for them to do so?
+
+Yes — pull requests are welcomed (see `CONTRIBUTING.md` and `CODE_OF_CONDUCT.md`). Contributions that change data semantics require a Croissant update and a Datasheet-section review.
+
+---
+
+## Lifecycle Statement (NeurIPS 2026 E&D track requirement)
+
+- **Status**: active.
+- **Versioning**: semantic; authoritative source is `CITATION.cff` `version` and `croissant.json` `version`.
+- **Maintenance owner**: Anonymous Group open-source research SIG; issues/PRs on GitHub.
+- **Deprecation policy**: prior minor version remains supported for at least one subsequent minor release; deprecations announced in `CHANGELOG.md` and the Hugging Face dataset README.
+- **Security contact**: see `SECURITY.md`.

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 ANONYMOUS
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -0,0 +1,554 @@
+---
+configs:
+- config_name: 64x64_samples
+  data_files:
+  - path: 64x64_samples/train/*
+    split: train
+  - path: 64x64_samples/calibration/*
+    split: calibration
+  - path: 64x64_samples/validation/*
+    split: validation
+  - path: 64x64_samples/test/*
+    split: test
+- config_name: 128x128_samples
+  data_files:
+  - path: 128x128_samples/train/*
+    split: train
+  - path: 128x128_samples/calibration/*
+    split: calibration
+  - path: 128x128_samples/validation/*
+    split: validation
+  - path: 128x128_samples/test/*
+    split: test
+- config_name: 256x256_samples
+  data_files:
+  - path: 256x256_samples/train/*
+    split: train
+  - path: 256x256_samples/calibration/*
+    split: calibration
+  - path: 256x256_samples/validation/*
+    split: validation
+  - path: 256x256_samples/test/*
+    split: test
+- config_name: 512x512_samples
+  data_files:
+  - path: 512x512_samples/train/*
+    split: train
+  - path: 512x512_samples/calibration/*
+    split: calibration
+  - path: 512x512_samples/validation/*
+    split: validation
+  - path: 512x512_samples/test/*
+    split: test
+- config_name: 1024x1024_samples
+  data_files:
+  - path: 1024x1024_samples/train/*
+    split: train
+  - path: 1024x1024_samples/calibration/*
+    split: calibration
+  - path: 1024x1024_samples/validation/*
+    split: validation
+  - path: 1024x1024_samples/test/*
+    split: test
+- config_name: 2048x2048_samples
+  data_files:
+  - path: 2048x2048_samples/train/*
+    split: train
+  - path: 2048x2048_samples/calibration/*
+    split: calibration
+  - path: 2048x2048_samples/validation/*
+    split: validation
+  - path: 2048x2048_samples/test/*
+    split: test
+- config_name: 4096x4096_samples
+  data_files:
+  - path: 4096x4096_samples/train/*
+    split: train
+  - path: 4096x4096_samples/calibration/*
+    split: calibration
+  - path: 4096x4096_samples/validation/*
+    split: validation
+  - path: 4096x4096_samples/test/*
+    split: test
+license: other
+language:
+- en
+task_categories:
+- other
+tags:
+- pathfinding
+- terrain-routing
+- digital-elevation-model
+- benchmark
+- remote-sensing
+- aw3d30
+pretty_name: PASSAGE (anonymous samples mirror)
+---
+
+<!-- PASSAGE Hugging Face Dataset Card (anonymous review mirror, samples only) -->
+
+> This is the **anonymous samples-only mirror** of PASSAGE for NeurIPS 2026 ED double-blind review.
+> The full release (all splits and full resolutions) is referenced from the paper's
+> `\anondata` macro and from the parent dataset card under the same anonymous handle.
+
+<!-- PASSAGE README — renders on both GitHub and Hugging Face -->
+
+<p align="center">
+  <img src="assets/banner.png" alt="PASSAGE — Path Solving Dataset & Generator" width="100%"/>
+</p>
+
+<h1 align="center">🌍 PASSAGE — Real-Terrain Benchmark & Generator</h1>
+
+<p align="center">
+  <strong>Trustworthiness-oriented real-terrain benchmark and generation pipeline for pathfinding and surrogate modeling.</strong>
+</p>
+
+<!-- Badges -->
+<p align="center">
+  <a href="https://github.com/anonymous-org/passage/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/anonymous-org/passage/ci.yml?label=CI&logo=githubactions&logoColor=white" alt="CI"></a>
+  <a href="https://github.com/anonymous-org/passage/actions/workflows/docs.yml"><img src="https://img.shields.io/github/actions/workflow/status/anonymous-org/passage/docs.yml?label=docs&logo=materialformkdocs&logoColor=white" alt="Docs"></a>
+  <a href="https://www.python.org/"><img src="https://img.shields.io/badge/python-≥3.11-3776AB?logo=python&logoColor=white" alt="Python"></a>
+  <a href="DATASET_LICENSE.md"><img src="https://img.shields.io/badge/dataset%20terms-AW3D30%20%2B%20MIT%20code-orange?logo=opensourceinitiative&logoColor=white" alt="Dataset Terms"></a>
+  <a href="https://huggingface.co/datasets/Anonym-2045/passage"><img src="https://img.shields.io/badge/%F0%9F%A4%97-Dataset-yellow" alt="HuggingFace Dataset"></a>
+  <a href="https://docs.astral.sh/ruff/"><img src="https://img.shields.io/badge/code%20style-ruff-000000?logo=ruff&logoColor=white" alt="Ruff"></a>
+  <a href="https://docs.astral.sh/uv/"><img src="https://img.shields.io/badge/package-uv-DE5FE9?logo=uv&logoColor=white" alt="uv"></a>
+  <a href="https://pre-commit.com/"><img src="https://img.shields.io/badge/pre--commit-enabled-brightgreen?logo=pre-commit&logoColor=white" alt="Pre-commit"></a>
+</p>
+
+<p align="center">
+  <a href="https://anonymous-org.github.io/passage/">📖 Documentation</a> •
+  <a href="https://huggingface.co/datasets/Anonym-2045/passage">🤗 Dataset</a> •
+  <a href="https://github.com/anonymous-org/passage">💻 GitHub</a> •
+  <a href="#-citation">📝 Cite</a>
+</p>
+
+---
+
+## Table of Contents
+
+- [Table of Contents](#table-of-contents)
+- [Get started](#get-started)
+- [Documentation](#documentation)
+- [🚁 Why PASSAGE?](#-why-passage)
+- [✨ Highlights](#-highlights)
+- [🗺️ Visual Summary](#️-visual-summary)
+- [🚀 Quick Start](#-quick-start)
+- [🔁 Pipeline Overview](#-pipeline-overview)
+- [🧭 Reproducibility](#-reproducibility)
+- [📦 Loading the Dataset](#-loading-the-dataset)
+- [⚙️ Configuration](#️-configuration)
+- [📓 Notebooks](#-notebooks)
+- [🔬 Research Applications](#-research-applications)
+- [📁 Output Structure](#-output-structure)
+- [🧪 Development \& Testing](#-development--testing)
+- [📝 Citation](#-citation)
+- [Contributing](#contributing)
+- [Security](#security)
+- [💬 Support](#-support)
+- [👤 Maintainers](#-maintainers)
+- [🙏 Acknowledgments](#-acknowledgments)
+- [License](#license)
+
+---
+
+## Get started
+
+PASSAGE is an open-source pathfinding dataset and generation toolkit built on real-world elevation data to benchmark and improve terrain-aware routing algorithms.
+
+- **Project goals**: provide reproducible, multi-resolution benchmarks for pathfinding research, ML surrogates, and safety-critical AI evaluation.
+- **Sponsoring SIG**: Anonymous Group open-source and critical AI research contributors.
+- **Community contact**: use [GitHub Issues](https://github.com/anonymous-org/passage/issues), [GitHub Discussions](https://github.com/anonymous-org/passage/discussions), or email `oss@anonymous.invalid`.
+
+For installation and first run commands, see [🚀 Quick Start](#-quick-start).
+
+## Documentation
+
+Documentation is available at [anonymous-org.github.io/passage](https://anonymous-org.github.io/passage/).
+
+The documentation site is published with GitHub Pages and includes user guides, API references, and architecture notes.
+
+Please also set the repository **About** metadata:
+- **Description**: short project purpose statement for PASSAGE.
+- **Website**: `https://anonymous-org.github.io/passage/`.
+
+---
+
+## 🚁 Why PASSAGE?
+
+> *Every year, Helicopter Emergency Medical Services (HEMS) crews across Europe perform over 300,000 rescue missions — racing against time through mountain passes, across valleys, and over ridgelines to reach patients in cardiac arrest or severe trauma. For every minute of delay, survival rates drop by 7–10%. The path the helicopter takes matters — yet pathfinding algorithms are still benchmarked on flat synthetic grids.*
+
+**PASSAGE** was created to change that. Built from **real-world satellite elevation data** (JAXA ALOS AW3D30 at 30 m/pixel), it provides the research community with large-scale, multi-resolution pathfinding benchmarks that capture the true complexity of terrain navigation.
+
+We open-source this toolkit to make terrain-routing methods easier to compare, reproduce, and audit under a shared benchmark surface.
+
+---
+
+## ✨ Highlights
+
+<table>
+<tr>
+  <td align="center" width="25%">🌐<br><strong>Multi-Resolution</strong><br><small>64×64 → 4096×4096 px</small></td>
+  <td align="center" width="25%">🏔️<br><strong>Real Elevation</strong><br><small>JAXA ALOS AW3D30 DSM</small></td>
+  <td align="center" width="25%">🔷<br><strong>Rich Obstacles</strong><br><small>Superellipse shapes</small></td>
+  <td align="center" width="25%">⚡<br><strong>3 A* Cost Variants</strong><br><small><code>elevation</code>, <code>slope_uphill</code>, <code>energy</code></small></td>
+</tr>
+<tr>
+  <td align="center">🧮<br><strong>Grid A* Oracle</strong><br><small>Exact labels + timings</small></td>
+  <td align="center">🤗<br><strong>HuggingFace-Ready</strong><br><small>Parquet + one-command upload</small></td>
+  <td align="center">🔁<br><strong>Fully Resumable</strong><br><small>No wasted compute</small></td>
+  <td align="center">✅<br><strong>Tested & CI/CD</strong><br><small>80%+ coverage, pre-commit</small></td>
+</tr>
+</table>
+
+<table>
+<tr>
+  <td align="center"><strong>4.4 M</strong><br><small>Default Samples</small></td>
+  <td align="center"><strong>7</strong><br><small>Resolution Levels</small></td>
+  <td align="center"><strong>3</strong><br><small>A* Configurations</small></td>
+  <td align="center"><strong>1</strong><br><small>Solver Backend</small></td>
+  <td align="center"><strong>30 m/px</strong><br><small>Spatial Resolution</small></td>
+</tr>
+</table>
+
+---
+
+## 🗺️ Visual Summary
+
+<p align="center">
+  <img src="assets/sample.png" alt="Representative PASSAGE sample" width="680"/>
+</p>
+
+<p align="center"><em>Normalized elevation with start/end markers, procedural superellipse obstacles, and two computed paths (free terrain vs. obstacle-aware).</em></p>
+
+---
+
+## 🚀 Quick Start
+
+```bash
+# Install uv (if needed)
+curl -LsSf https://astral.sh/uv/install.sh | sh
+
+# Clone & install
+git clone https://github.com/anonymous-org/passage.git
+cd passage && uv sync
+
+# One-command local rebuild (download -> calibrate -> generate -> export -> manifest)
+make rebuild RESOLUTION=256
+
+# Or run each stage manually
+passage download                       # 1️⃣  Fetch JAXA elevation tiles
+passage calibrate                      # 2️⃣  Compute global elevation stats
+passage generate --resolution 256      # 3️⃣  Generate pathfinding samples
+passage export --no-upload             # 4️⃣  Package as Parquet shards
+make manifest                          # 5️⃣  Write outputs/export/manifest.sha256
+
+# Point generate/export at a specific calibration file when needed
+passage generate --calibration-filepath /path/to/calibrate.json --resolution 256
+passage export --calibration-filepath /path/to/calibrate.json --no-upload
+```
+
+<details>
+<summary>💡 Alternative: install with pip</summary>
+
+```bash
+python -m venv .venv && source .venv/bin/activate
+pip install -e ".[dev]"
+```
+</details>
+
+---
+
+## 🔁 Pipeline Overview
+
+<table>
+<tr>
+  <td align="center"><strong>1. Download</strong><br><small>JAXA DSM tiles</small></td>
+  <td align="center"><strong>→</strong></td>
+  <td align="center"><strong>2. Calibrate</strong><br><small>min / max elevation</small></td>
+  <td align="center"><strong>→</strong></td>
+  <td align="center"><strong>3. Generate</strong><br><small>samples + paths</small></td>
+  <td align="center"><strong>→</strong></td>
+  <td align="center"><strong>4. Export</strong><br><small>Parquet + HF Hub</small></td>
+</tr>
+</table>
+
+| Step | Command | What it does |
+|:----:|---------|--------------|
+| 1 | `passage download` | Download JAXA ALOS AW3D30 5°×5° tile archives with resume |
+| 2 | `passage calibrate` | Scan tiles for global min/max elevation |
+| 3 | `passage generate` | Create multi-resolution samples with paths, obstacles, tensors |
+| 4 | `passage export` | Write Parquet shards, execute notebooks, push to HF Hub |
+
+> 📖 **Full walkthrough:** see the [User Guide](https://anonymous-org.github.io/passage/guide/) or the local [USAGE.md](USAGE.md) reference.
+
+---
+
+## 🧭 Reproducibility
+
+- `splits.seed` controls deterministic split assignment.
+- `generation.seed` controls deterministic per-sample tile selection, crop placement, marker placement, and obstacle synthesis.
+- Each sample metadata file records `provenance.generation_seed` and `provenance.sample_seed`.
+- Runtime timestamps remain informational metadata and are not part of the deterministic sampling contract.
+- `make manifest` writes `outputs/export/manifest.sha256` so a released export tree can be checked for completeness.
+
+Repository-root release metadata is tracked in [`CITATION.cff`](CITATION.cff), [`DATASHEET.md`](DATASHEET.md), [`CHANGELOG.md`](CHANGELOG.md), and [`croissant.json`](croissant.json).
+
+---
+
+## 📦 Loading the Dataset
+
+```python
+from datasets import load_dataset
+from passage.utils import decode_tensor_blob
+
+# Load a specific resolution from Hugging Face
+ds = load_dataset("Anonym-2045/passage", name="256x256")
+sample = ds["train"][0]
+
+# Decode the compressed tensor → (256, 256, 3) numpy array
+tensor = decode_tensor_blob(sample["tensor"], (256, 256, 3))
+elevation = tensor[:, :, 0]  # Normalized [0, 1]
+markers   = tensor[:, :, 1]  # -1 background, 0 start, 1 goal
+obstacles = tensor[:, :, 2]  # Binary mask (0/1)
+
+# Access solver paths
+path_cols = [k for k in sample if k.startswith("path_")]
+print("Solver paths:", path_cols)
+```
+
+---
+
+## ⚙️ Configuration
+
+Core settings live in [`config/config.yaml`](config/config.yaml):
+
+```yaml
+resolutions: [64, 128, 256, 512, 1024, 2048, 4096]
+num_samples:
+  64:    1000000
+  128:   1000000
+  256:   1000000
+  512:   1000000
+  1024:   250000
+  2048:   100000
+  4096:    50000
+splits:
+  ratios: {train: 0.90, calibration: 0.02, validation: 0.04, test: 0.04}
+  seed: 42
+generation:
+  seed: 42
+  preview_images:
+    enabled: true
+    resolutions: [64, 128, 256, 512, 1024, 2048, 4096]
+    num_samples_per_split: 64
+obstacles:
+  enabled: true
+  ratio_max: 0.30
+  shape:
+    n: { distribution: log_uniform, min: 1.0, max: 80.0 }
+solvers:
+  - name: astar_elevation
+    backend: grid
+    solver: astar
+    cost_model: elevation
+    weight: 5878.0
+```
+
+CLI options override config values. See the [Configuration Reference](https://anonymous-org.github.io/passage/guide/configuration/) in the documentation.
+
+<details>
+<summary>🔐 Hugging Face credentials</summary>
+
+```bash
+# .env
+HF_USER_NAME=your-hf-username
+HF_REPO_NAME=passage-dataset
+
+# .env.secret (never commit!)
+HF_TOKEN=hf_XXXXXXXXXXXXXXXXXXXXXXXX
+```
+</details>
+
+---
+
+## 📓 Notebooks
+
+| Notebook | Description |
+|----------|-------------|
+| [`demo.ipynb`](notebooks/demo.ipynb) | Quick-start: load, decode, and visualize a sample |
+| [`costmodel.ipynb`](notebooks/costmodel.ipynb) | Compare the configured terrain-cost variants on the same terrain |
+| [`obstacles.ipynb`](notebooks/obstacles.ipynb) | Explore superellipse obstacle generation |
+| [`solve.ipynb`](notebooks/solve.ipynb) | Solver timing and quality comparison |
+| `_dataset.ipynb` | Template — auto-executed per resolution during export |
+
+---
+
+## 🔬 Research Applications
+
+| Domain | Use Case |
+|--------|----------|
+| **Classical AI** | A\*, Dijkstra benchmarking across cost models |
+| **Reinforcement Learning** | Grid-based terrain navigation agents |
+| **Graph Neural Networks** | Learned heuristics on elevation graphs |
+| **Surrogate Models** | Fast path prediction from 4.4M+ multi-resolution terrain-routing examples |
+| **Computer Vision** | Path segmentation from elevation images |
+| **Certified ML** | ARP 6983 / ED-324 — bounded ODD, deterministic data |
+
+---
+
+## 📁 Output Structure
+
+```
+outputs/
+├── download/           # Tile archives + download.csv + extracted GeoTIFFs
+├── calibrate/          # calibrate.json + elevation maps
+├── dataset/            # Per-resolution samples
+│   └── 256/
+│       └── train/
+│           ├── tensors/N045/E010/0000000000.npy.zst
+│           ├── metadata/N045/E010/0000000000.yaml
+│           ├── images/astar_energy/N045/E010/0000000000.png
+│           └── paths/astar_energy/free/N045/E010/0000000000.csv
+└── export/             # Hugging Face-ready parquet shards + release checksums
+  ├── 256x256/train/data-00000.parquet
+  ├── 256x256_samples/train/data-00000.parquet
+  ├── README.md
+  └── manifest.sha256
+```
+
+---
+
+## 🧪 Development & Testing
+
+```bash
+# Install dev dependencies
+uv sync --group dev
+
+# Run tests (80%+ coverage required)
+uv run pytest --cov
+
+# Lint & format
+uv run ruff check src/ tests/
+uv run ruff format src/ tests/
+
+# Pre-commit hooks
+uv run pre-commit install
+uv run pre-commit run --all-files
+```
+
+---
+
+## 📝 Citation
+
+Citation metadata is available in [`CITATION.cff`](CITATION.cff). If you prefer BibTeX, use:
+
+```bibtex
+@inproceedings{anonymous2026passage,
+  title     = {{PASSAGE}: A Real-Terrain Multi-Resolution Benchmark for
+               Trustworthy Constrained Path Planning},
+  author    = {Anonymous, Author and Anonymous, Anonymous and Anonymous, Author},
+  booktitle = {Advances in Neural Information Processing Systems (NeurIPS),
+               Datasets and Benchmarks Track},
+  year      = {2026},
+  url       = {https://github.com/anonymous-org/passage}
+}
+```
+
+---
+
+## Contributing
+
+If you are interested in contributing to PASSAGE, start by reading the [Contributing guide](/CONTRIBUTING.md).
+
+Contributions are welcome; the guide covers development setup, coding standards, pull requests, and issue reporting.
+
+In brief:
+
+```bash
+# Fork & clone, then:
+uv sync --group dev
+uv run pre-commit install
+git checkout -b feature/your-feature-name
+# ... make changes ...
+uv run pytest --cov
+uv run ruff check src/ tests/
+```
+
+---
+
+## 💬 Support
+
+Need help? Here's where to go:
+
+| Channel | Link |
+|---------|------|
+| **Documentation** | [anonymous-org.github.io/passage](https://anonymous-org.github.io/passage/) |
+| **Bug Reports** | [GitHub Issues](https://github.com/anonymous-org/passage/issues) |
+| **Discussions & Q&A** | [GitHub Discussions](https://github.com/anonymous-org/passage/discussions) |
+| **Email** | anonymous@anonymous.invalid |
+
+Before opening an issue, please search existing issues and discussions to avoid duplicates.
+
+---
+
+## 👤 Maintainers
+
+| Name | Affiliation | Contact |
+|------|------------|--------|
+| **Anonymous Author** | [Anonymous Group](https://www.anonymous-org.com/) | [anonymous@anonymous.invalid](mailto:anonymous@anonymous.invalid) |
+| **Anonymous Author** | [Anonymous Group](https://www.anonymous-org.com/) | [anonymous@anonymous.invalid](mailto:anonymous@anonymous.invalid) |
+
+---
+
+## 🙏 Acknowledgments
+
+- **[JAXA](https://www.eorc.jaxa.jp/ALOS/en/dataset/aw3d30/aw3d30_e.htm)** for the ALOS AW3D30 Digital Surface Model
+- **[Hugging Face](https://huggingface.co/)** for the Datasets library and Hub infrastructure
+- **[Anonymous Group](https://www.anonymous-org.com/)** for supporting open-source research in safety-critical AI
+
+---
+
+## Security
+
+Security reporting and handling procedures are documented in [SECURITY.md](SECURITY.md).
+
+If you discover a vulnerability, please use private reporting channels described in that file.
+
+---
+
+## License
+
+PASSAGE ships **two distinct licensing surfaces** and you should
+recognise which one applies to what you are doing:
+
+| Surface | License | What it covers |
+| --- | --- | --- |
+| Generator and evaluator code | [MIT License](LICENSE) | every Python file in this repository — the dataset generator (`passage`), the evaluation contract (`passage-eval`), and the documentation source. |
+| Generated elevation data | [JAXA AW3D30 terms of use](https://www.eorc.jaxa.jp/ALOS/en/dataset/aw3d30/aw3d30_e.htm) | every elevation tile that PASSAGE ingests, every sample whose elevation channel is derived from those tiles, and any model whose training data carries that channel. The full text mirrored at release time is in [`DATASET_LICENSE.md`](DATASET_LICENSE.md). |
+
+> [!WARNING]
+> **Respect JAXA's licensing terms.** PASSAGE is built on the
+> [JAXA ALOS World 3D 30 m DSM (AW3D30)](https://www.eorc.jaxa.jp/ALOS/en/dataset/aw3d30/aw3d30_e.htm).
+> Any user of PASSAGE — or of any artefact derived from it — **must
+> respect the JAXA terms of use**:
+>
+> 1. **Cite the JAXA AW3D30 source** in publications using PASSAGE
+>    (recommended: Takaku *et al.*, *"Updates of 'AW3D30' ALOS Global
+>    DSM with Other Open Access Datasets,"* ISPRS Archives, 2020).
+> 2. **Reproduce the AW3D30 attribution** in any redistribution of
+>    the elevation channel or of any artefact directly derived from
+>    it.
+> 3. **Read and abide by the JAXA terms-of-use page** linked above
+>    for the authoritative wording.
+>
+> We sincerely thank JAXA for openly sharing AW3D30 with the research
+> community.
+
+---
+
+<p align="center">
+  <a href="https://anonymous-org.github.io/passage/">📖 Documentation</a> •
+  <a href="https://huggingface.co/datasets/Anonym-2045/passage">🤗 Dataset</a> •
+  <a href="https://github.com/anonymous-org/passage/issues">🐛 Issues</a> •
+  <a href="https://github.com/anonymous-org/passage/discussions">💬 Discussions</a>
+</p>

croissant.json

@@ -0,0 +1,576 @@
+{
+  "@context": {
+    "@language": "en",
+    "@vocab": "https://schema.org/",
+    "citeAs": "cr:citeAs",
+    "column": "cr:column",
+    "conformsTo": "dct:conformsTo",
+    "containedIn": "cr:containedIn",
+    "cr": "http://mlcommons.org/croissant/",
+    "prov": "http://www.w3.org/ns/prov#",
+    "rai": "http://mlcommons.org/croissant/RAI/",
+    "data": {
+      "@id": "cr:data",
+      "@type": "@json"
+    },
+    "dataType": {
+      "@id": "cr:dataType",
+      "@type": "@vocab"
+    },
+    "dct": "http://purl.org/dc/terms/",
+    "equivalentProperty": "cr:equivalentProperty",
+    "examples": {
+      "@id": "cr:examples",
+      "@type": "@json"
+    },
+    "extract": "cr:extract",
+    "field": "cr:field",
+    "fileProperty": "cr:fileProperty",
+    "fileObject": "cr:fileObject",
+    "fileSet": "cr:fileSet",
+    "format": "cr:format",
+    "includes": "cr:includes",
+    "isLiveDataset": "cr:isLiveDataset",
+    "jsonPath": "cr:jsonPath",
+    "key": "cr:key",
+    "md5": "cr:md5",
+    "parentField": "cr:parentField",
+    "path": "cr:path",
+    "recordSet": "cr:recordSet",
+    "references": "cr:references",
+    "regex": "cr:regex",
+    "repeated": "cr:repeated",
+    "replace": "cr:replace",
+    "samplingRate": "cr:samplingRate",
+    "sc": "https://schema.org/",
+    "separator": "cr:separator",
+    "source": "cr:source",
+    "subField": "cr:subField",
+    "transform": "cr:transform"
+  },
+  "@id": "https://huggingface.co/datasets/Anonym-2045/passage",
+  "@type": "sc:Dataset",
+  "name": "PASSAGE",
+  "description": "PASSAGE is a multi-resolution terrain-routing benchmark generated from JAXA ALOS AW3D30 elevation data. The default benchmark exports 4,400,000 samples (1,000,000 per resolution at 64x64, 128x128, 256x256, 512x512; 250,000 at 1024x1024; 100,000 at 2048x2048; 50,000 at 4096x4096), with deterministic 90/2/4/4 train/calibration/validation/test split assignment and deterministic per-sample seeding for tile selection, crop placement, marker placement, and obstacle synthesis.",
+  "url": "https://huggingface.co/datasets/Anonym-2045/passage",
+  "version": "0.1.0",
+  "license": "https://github.com/anonymous-org/passage/blob/main/DATASET_LICENSE.md",
+  "keywords": [
+    "pathfinding",
+    "terrain routing",
+    "digital elevation model",
+    "benchmark",
+    "remote sensing"
+  ],
+  "datePublished": "2026-03-30",
+  "creator": [
+    {
+      "@type": "sc:Person",
+      "name": "Anonymous Author"
+    },
+    {
+      "@type": "sc:Person",
+      "name": "Anonymous Author"
+    },
+    {
+      "@type": "sc:Person",
+      "name": "Anonymous Author"
+    }
+  ],
+  "conformsTo": "http://mlcommons.org/croissant/1.1",
+  "citeAs": "@inproceedings{passage2026,\n  title     = {PASSAGE: A Real-Terrain Multi-Resolution Benchmark for Constrained Path Planning},\n  author    = {Anonymous, Author and Anonymous, Author and Anonymous, Author},\n  booktitle = {Advances in Neural Information Processing Systems},\n  year      = {2026},\n  note      = {Evaluations \\& Datasets Track (submission)}\n}",
+  "distribution": [
+    {
+      "@id": "hf-passage-tree",
+      "@type": "cr:FileObject",
+      "name": "Hugging Face PASSAGE dataset repository tree",
+      "description": "The Hugging Face git repository hosting the PASSAGE parquet shards.",
+      "contentUrl": "https://huggingface.co/datasets/Anonym-2045/passage",
+      "encodingFormat": "git+https",
+      "sha256": "https://github.com/mlcommons/croissant/issues/80"
+    },
+    {
+      "@id": "passage-full-parquet",
+      "@type": "cr:FileSet",
+      "name": "PASSAGE full parquet shards",
+      "description": "All full PASSAGE parquet configs from 64x64 through 4096x4096.",
+      "containedIn": {
+        "@id": "hf-passage-tree"
+      },
+      "encodingFormat": "application/x-parquet",
+      "includes": [
+        "64x64/*/*.parquet",
+        "128x128/*/*.parquet",
+        "256x256/*/*.parquet",
+        "512x512/*/*.parquet",
+        "1024x1024/*/*.parquet",
+        "2048x2048/*/*.parquet",
+        "4096x4096/*/*.parquet"
+      ]
+    },
+    {
+      "@id": "passage-sample-parquet",
+      "@type": "cr:FileSet",
+      "name": "PASSAGE sample parquet shards",
+      "description": "Sample PASSAGE parquet configs with per-solver preview images.",
+      "containedIn": {
+        "@id": "hf-passage-tree"
+      },
+      "encodingFormat": "application/x-parquet",
+      "includes": [
+        "64x64_samples/*/*.parquet",
+        "128x128_samples/*/*.parquet",
+        "256x256_samples/*/*.parquet",
+        "512x512_samples/*/*.parquet",
+        "1024x1024_samples/*/*.parquet",
+        "2048x2048_samples/*/*.parquet",
+        "4096x4096_samples/*/*.parquet"
+      ]
+    }
+  ],
+  "recordSet": [
+    {
+      "@id": "passage-full-records",
+      "@type": "cr:RecordSet",
+      "name": "PASSAGE full parquet records",
+      "description": "Rows from the non-sample PASSAGE parquet exports.",
+      "field": [
+        {
+          "@id": "passage-full/id",
+          "@type": "cr:Field",
+          "name": "id",
+          "description": "Zero-padded sample identifier.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-full-parquet"
+            },
+            "extract": {
+              "column": "id"
+            }
+          }
+        },
+        {
+          "@id": "passage-full/tensor",
+          "@type": "cr:Field",
+          "name": "tensor",
+          "description": "Compressed binary tensor storing normalized elevation, markers, and obstacles.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-full-parquet"
+            },
+            "extract": {
+              "column": "tensor"
+            }
+          }
+        },
+        {
+          "@id": "passage-full/path-astar-elevation-free",
+          "@type": "cr:Field",
+          "name": "path_astar_elevation_free",
+          "description": "Pixel path for the free-terrain astar_elevation annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-full-parquet"
+            },
+            "extract": {
+              "column": "path_astar_elevation_free"
+            }
+          }
+        },
+        {
+          "@id": "passage-full/path-astar-elevation-obstacles",
+          "@type": "cr:Field",
+          "name": "path_astar_elevation_obstacles",
+          "description": "Pixel path for the obstacle-aware astar_elevation annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-full-parquet"
+            },
+            "extract": {
+              "column": "path_astar_elevation_obstacles"
+            }
+          }
+        },
+        {
+          "@id": "passage-full/path-astar-energy-free",
+          "@type": "cr:Field",
+          "name": "path_astar_energy_free",
+          "description": "Pixel path for the free-terrain astar_energy annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-full-parquet"
+            },
+            "extract": {
+              "column": "path_astar_energy_free"
+            }
+          }
+        },
+        {
+          "@id": "passage-full/path-astar-energy-obstacles",
+          "@type": "cr:Field",
+          "name": "path_astar_energy_obstacles",
+          "description": "Pixel path for the obstacle-aware astar_energy annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-full-parquet"
+            },
+            "extract": {
+              "column": "path_astar_energy_obstacles"
+            }
+          }
+        },
+        {
+          "@id": "passage-full/path-astar-slope-free",
+          "@type": "cr:Field",
+          "name": "path_astar_slope_free",
+          "description": "Pixel path for the free-terrain astar_slope annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-full-parquet"
+            },
+            "extract": {
+              "column": "path_astar_slope_free"
+            }
+          }
+        },
+        {
+          "@id": "passage-full/path-astar-slope-obstacles",
+          "@type": "cr:Field",
+          "name": "path_astar_slope_obstacles",
+          "description": "Pixel path for the obstacle-aware astar_slope annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-full-parquet"
+            },
+            "extract": {
+              "column": "path_astar_slope_obstacles"
+            }
+          }
+        },
+        {
+          "@id": "passage-full/metadata",
+          "@type": "cr:Field",
+          "name": "metadata",
+          "description": "Structured sample metadata including split, resolution, crop geometry, solver settings, timing, and deterministic provenance.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-full-parquet"
+            },
+            "extract": {
+              "column": "metadata"
+            }
+          }
+        }
+      ]
+    },
+    {
+      "@id": "passage-sample-records",
+      "@type": "cr:RecordSet",
+      "name": "PASSAGE sample parquet records",
+      "description": "Rows from the PASSAGE sample parquet exports with preview images.",
+      "field": [
+        {
+          "@id": "passage-sample/id",
+          "@type": "cr:Field",
+          "name": "id",
+          "description": "Zero-padded sample identifier.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "id"
+            }
+          }
+        },
+        {
+          "@id": "passage-sample/image-astar-elevation",
+          "@type": "cr:Field",
+          "name": "image_astar_elevation",
+          "description": "Preview image rendered for the astar_elevation solver.",
+          "dataType": "sc:ImageObject",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "image_astar_elevation"
+            }
+          }
+        },
+        {
+          "@id": "passage-sample/image-astar-energy",
+          "@type": "cr:Field",
+          "name": "image_astar_energy",
+          "description": "Preview image rendered for the astar_energy solver.",
+          "dataType": "sc:ImageObject",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "image_astar_energy"
+            }
+          }
+        },
+        {
+          "@id": "passage-sample/image-astar-slope",
+          "@type": "cr:Field",
+          "name": "image_astar_slope",
+          "description": "Preview image rendered for the astar_slope solver.",
+          "dataType": "sc:ImageObject",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "image_astar_slope"
+            }
+          }
+        },
+        {
+          "@id": "passage-sample/tensor",
+          "@type": "cr:Field",
+          "name": "tensor",
+          "description": "Compressed binary tensor storing normalized elevation, markers, and obstacles.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "tensor"
+            }
+          }
+        },
+        {
+          "@id": "passage-sample/path-astar-elevation-free",
+          "@type": "cr:Field",
+          "name": "path_astar_elevation_free",
+          "description": "Pixel path for the free-terrain astar_elevation annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "path_astar_elevation_free"
+            }
+          }
+        },
+        {
+          "@id": "passage-sample/path-astar-elevation-obstacles",
+          "@type": "cr:Field",
+          "name": "path_astar_elevation_obstacles",
+          "description": "Pixel path for the obstacle-aware astar_elevation annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "path_astar_elevation_obstacles"
+            }
+          }
+        },
+        {
+          "@id": "passage-sample/path-astar-energy-free",
+          "@type": "cr:Field",
+          "name": "path_astar_energy_free",
+          "description": "Pixel path for the free-terrain astar_energy annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "path_astar_energy_free"
+            }
+          }
+        },
+        {
+          "@id": "passage-sample/path-astar-energy-obstacles",
+          "@type": "cr:Field",
+          "name": "path_astar_energy_obstacles",
+          "description": "Pixel path for the obstacle-aware astar_energy annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "path_astar_energy_obstacles"
+            }
+          }
+        },
+        {
+          "@id": "passage-sample/path-astar-slope-free",
+          "@type": "cr:Field",
+          "name": "path_astar_slope_free",
+          "description": "Pixel path for the free-terrain astar_slope annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "path_astar_slope_free"
+            }
+          }
+        },
+        {
+          "@id": "passage-sample/path-astar-slope-obstacles",
+          "@type": "cr:Field",
+          "name": "path_astar_slope_obstacles",
+          "description": "Pixel path for the obstacle-aware astar_slope annotation.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "path_astar_slope_obstacles"
+            }
+          }
+        },
+        {
+          "@id": "passage-sample/metadata",
+          "@type": "cr:Field",
+          "name": "metadata",
+          "description": "Structured sample metadata including split, resolution, crop geometry, solver settings, timing, and deterministic provenance.",
+          "dataType": "sc:Text",
+          "source": {
+            "fileSet": {
+              "@id": "passage-sample-parquet"
+            },
+            "extract": {
+              "column": "metadata"
+            }
+          }
+        }
+      ]
+    }
+  ],
+  "rai:dataCollection": "Observational JAXA ALOS AW3D30 digital surface model tiles combined with deterministic procedural obstacle synthesis and solver-generated path annotations.",
+  "rai:dataCollectionType": [
+    "observational",
+    "synthetic",
+    "computational annotation"
+  ],
+  "rai:dataCollectionRawData": "Raw source tiles are downloaded from JAXA ALOS AW3D30 5x5 degree archives; PASSAGE generation operates on 1x1 degree 3600x3600 sub-tiles at approximately 30 meters per pixel.",
+  "rai:dataPreprocessingProtocol": [
+    "Calibrate global elevation minima and maxima over the local DSM cache.",
+    "Use deterministic per-sample seeding to choose a target elevation, tile, crop, and markers.",
+    "Optionally synthesize superellipse obstacles under the configured coverage and size limits.",
+    "Solve reference paths with the configured astar_elevation, astar_energy, and astar_slope solvers, with and without obstacles.",
+    "Export tensors, metadata, CSV paths, preview images, notebooks, and parquet shards."
+  ],
+  "rai:dataAnnotationProtocol": "Reference paths are computational annotations produced by the configured pathfinding solvers over the generated terrain crops and obstacle masks.",
+  "rai:dataReleaseMaintenancePlan": "Releases are versioned in CHANGELOG.md, regenerated through the repository Makefile, and verified with outputs/export/manifest.sha256.",
+  "rai:hasSyntheticData": true,
+  "prov:wasDerivedFrom": [
+    {
+      "@id": "https://www.eorc.jaxa.jp/ALOS/en/dataset/aw3d30/aw3d30_e.htm",
+      "prov:label": "JAXA ALOS Global Digital Surface Model 'ALOS World 3D - 30m' (AW3D30)",
+      "sc:description": "Global 30 m/pixel digital surface model derived from the PRISM panchromatic stereo sensor on the ALOS satellite. PASSAGE samples its terrain content directly from AW3D30 5x5 degree archive tiles, retaining the native 30 m/pixel resolution as the canonical observational layer; obstacles, start/goal markers, and reference paths are then synthesized on top.",
+      "sc:license": "https://earth.jaxa.jp/policy/en.html",
+      "prov:wasAttributedTo": {
+        "@id": "https://global.jaxa.jp/",
+        "prov:label": "Japan Aerospace Exploration Agency (JAXA)"
+      }
+    }
+  ],
+  "prov:wasGeneratedBy": [
+    {
+      "@type": "prov:Activity",
+      "prov:type": {"@id": "https://www.wikidata.org/wiki/Q4929239"},
+      "prov:label": "AW3D30 tile collection",
+      "sc:description": "Automated download of JAXA ALOS AW3D30 5x5 degree GeoTIFF archive tiles into the local DSM cache covering the configured PASSAGE Operational Design Domain (ODD). No manual sampling: the tile list is generated deterministically from the ODD specification in config/config.yaml and downloaded by the make data target.",
+      "prov:atLocation": "https://github.com/anonymous-org/passage",
+      "prov:wasAttributedTo": [
+        {
+          "@type": "prov:SoftwareAgent",
+          "@id": "https://github.com/anonymous-org/passage/tree/main/src/passage",
+          "prov:label": "PASSAGE generation pipeline (Python, Numba)",
+          "sc:description": "Open-source generator implemented in Python with Numba-JIT inner kernels. Deterministic, no human in the loop."
+        }
+      ]
+    },
+    {
+      "@type": "prov:Activity",
+      "prov:type": {"@id": "https://www.wikidata.org/wiki/Q5227332"},
+      "prov:label": "Sample preprocessing and procedural obstacle synthesis",
+      "sc:description": "Per-sample preprocessing: (i) calibrate global elevation minima and maxima from the local DSM cache; (ii) derive a per-sample blake2b seed (src/passage/seeds.py); (iii) select a 1x1 degree sub-tile and an in-tile crop; (iv) place start/goal markers respecting the configured minimum start-goal distance; (v) optionally synthesize super-ellipse obstacle masks. No human annotators or crowdworkers participate; every step is deterministic and reproducible from the seed and config.",
+      "prov:atLocation": "https://github.com/anonymous-org/passage",
+      "prov:wasAttributedTo": [
+        {
+          "@type": "prov:SoftwareAgent",
+          "@id": "https://github.com/anonymous-org/passage/tree/main/src/passage",
+          "prov:label": "PASSAGE generation pipeline (Python, Numba)"
+        }
+      ]
+    },
+    {
+      "@type": "prov:Activity",
+      "prov:type": {"@id": "https://www.wikidata.org/wiki/Q109719325"},
+      "prov:label": "Reference-path annotation by exact A* solvers",
+      "sc:description": "Computational (machine-only) annotation. For every (terrain, start, goal, obstacle) sample, exact reference paths are produced by the Numba-JIT grid A* in src/passage/pathfinding_utils.py under the three pre-registered cost models (astar_elevation, astar_energy, astar_slope), both with and without obstacles. No human annotators, no language-model agents, no crowdsourcing platform: annotations are exact under the documented cost model and validated by construction (same solver, same heuristic, same tie-break across all consumers of the dataset).",
+      "prov:atLocation": "https://github.com/anonymous-org/passage/blob/main/src/passage/pathfinding_utils.py",
+      "prov:wasAttributedTo": [
+        {
+          "@type": "prov:SoftwareAgent",
+          "@id": "https://github.com/anonymous-org/passage/blob/main/src/passage/pathfinding_utils.py",
+          "prov:label": "Numba-JIT grid-backend A* solver",
+          "sc:description": "Deterministic exact-shortest-path solver. Every annotation is reproducible from the per-sample blake2b seed, the cost model id, and the obstacle flag."
+        }
+      ]
+    },
+    {
+      "@type": "prov:Activity",
+      "prov:type": {"@id": "https://www.wikidata.org/wiki/Q3306762"},
+      "prov:label": "Manifest and shard quality review",
+      "sc:description": "Post-generation quality review. Every released shard is hashed and recorded in outputs/export/manifest.sha256; the make verify target re-checks shard hashes and counts against config/config.yaml before a release is tagged. No subjective review: the gate is exact-bytes equality with the manifest captured at generation time.",
+      "prov:atLocation": "https://github.com/anonymous-org/passage",
+      "prov:wasAttributedTo": [
+        {
+          "@type": "prov:SoftwareAgent",
+          "@id": "https://github.com/anonymous-org/passage/blob/main/Makefile",
+          "prov:label": "PASSAGE release-verification Makefile targets"
+        }
+      ]
+    }
+  ],
+  "rai:personalSensitiveInformation": "None. PASSAGE contains no personally identifiable information, no biometric data, and no human-subject content. Every sample is a synthetic terrain-routing instance built on top of the public JAXA ALOS AW3D30 elevation product; obstacles, start/goal markers, and reference paths are procedurally generated from a deterministic blake2b seed.",
+  "rai:dataBiases": [
+    "Coverage and elevation fidelity are limited by the underlying AW3D30 product and the chosen download footprint.",
+    "Terrain-only routing omits weather, airspace, vegetation, built structures, and operational constraints."
+  ],
+  "rai:dataLimitations": [
+    "Procedural obstacle masks are abstractions rather than measured hazards.",
+    "Deterministic generation covers sample content; runtime timestamps remain informational metadata."
+  ],
+  "rai:dataUseCases": [
+    "Benchmark terrain-aware path planners.",
+    "Train and evaluate learned routing surrogates in vision, graph, and reinforcement-learning settings.",
+    "Study reproducibility and geographic hold-out generalization across resolutions."
+  ],
+  "rai:dataSocialImpact": "PASSAGE is intended for research and evaluation of routing methods in safety-relevant terrain settings. It is not a certified operational navigation product.",
+  "rai:annotationsPerItem": "Every sample carries exactly one ordered path annotation per configured cost model (default: astar_elevation, astar_energy, astar_slope) and, where applicable, per obstacle variant (free-terrain and obstacle-aware). No multi-annotator redundancy.",
+  "rai:annotatorDemographics": "Not applicable: annotations are produced exclusively by the Numba-JIT A* grid solver documented in src/passage/pathfinding_utils.py. No human annotators were involved.",
+  "rai:dataAnnotationAnalysis": "Annotations are exact under the documented cost model and are validated by construction (same solver, same heuristic, same connectivity). Consistency is enforced by deterministic seeding and by the outputs/export/manifest.sha256 checksum file bundled with each release.",
+  "rai:dataAnnotationPlatform": "None: annotations are computational. Generation is driven by the Makefile targets in this repository, executed on CPU with optional GPU used only by downstream baselines, not by dataset generation.",
+  "rai:machineAnnotationTools": [
+    "Numba-JIT grid-backend A* implemented in src/passage/pathfinding_utils.py",
+    "Deterministic blake2b per-sample seed derivation documented in src/passage/seeds.py"
+  ]
+}