Published June 19, 2026 | Version v0.0.1
Dataset Open

When Active Learning Fails, Uncalibrated Out of Distribution Uncertainty Quantification Might Be the Problem Data and Model Artifacts

Authors/Creators

  • 1. ROR icon King Abdullah University of Science and Technology
  • 2. ROR icon Carnegie Mellon University
  • 3. ROR icon University of Toronto

Description

Model checkpoints and result artifacts supporting the paper \"When Active Learning Fails, Uncalibrated Out of Distribution Uncertainty Quantification Might Be the Problem\" (arXiv:2511.17760). Includes ALIGNN ensemble checkpoints, loss-landscape sampling results, active-learning experiment histories, and the featurized JARVIS-22 dataset used for training.

Files

_curvature_study_output_alignn.zip

Files (25.3 GB)

Name Size
md5:aa3761f5ebe87533111b8812b65aadf5
7.5 MB Preview Download
md5:e6409e2d5148d07a6697067a38102409
158.1 MB Preview Download
md5:35bc0a21f18413cb94b0f51043368952
2.9 GB Preview Download
md5:a9d9fdfa855c9f935e0ab07e1db5ee49
2.9 GB Preview Download
md5:6e83f26331cbd109388cf557170fcd99
2.9 GB Preview Download
md5:fbf0bfe970fde6c7227fbef7488bd4a1
2.9 GB Preview Download
md5:de14462c753d43644e1ff9c62f267726
215.8 MB Preview Download
md5:6b075f57bcf22f79714b82fe64282e3a
1.4 GB Preview Download
md5:6d4db1a64d2ab1c7b47bdf332d20e473
92.1 MB Preview Download
md5:a7eb174fdef2478c2ad4e08ff7e89b9c
322.3 MB Preview Download
md5:ce7554533232369c667c8febeb44cd77
77.2 MB Preview Download
md5:f70eeb4d4d245b22309b1f0d8d0b22b4
7.5 MB Preview Download
md5:f582c8b1d9e05173e599714b9bb0d60a
2.9 GB Preview Download
md5:13618666de839cc7bb6521ac7dfaa932
2.9 GB Preview Download
md5:6389a5806d10c1ef43e35041fee6c086
2.9 GB Preview Download
md5:f1d83e04657fb9dafff15934dc412df1
2.0 GB Preview Download
md5:f6fd5194762b7851c7b04be0f6561825
3.8 MB Preview Download
md5:6bf3b6bdf84594e8f687fbdd09f47a5c
780.6 MB Preview Download

Additional details

Dates

Submitted
2026-06-21

Software

Repository URL
https://github.com/TRustworthy-AI-Tools-for-Science/paper_when_active_learning_fails