Published June 27, 2025 | Version v1

Dataset and results for dPL and MdPL Experiments

Authors/Creators

  • 1. Institute of Industrial Science, University of Tokyo

Description

This deposition contains compressed archives of all major model outputs and training artifacts for differentiable Parameter Learning (dPL) and Multi-task dPL (MdPL) experiments. Included are: - dPL_Gz: dPL Gaussian-zipped models and logs - dPL_Surrogate: dPL surrogate model weights and logs - ILS_dPL: ILS simulation outputs calibrated with dPL - ILS_LOOCV: Leave-one-out CV results for ILS - ILS_MdPL: ILS outputs using Multi-task dPL calibration - ILS_ORI: Original uncalibrated ILS outputs - MdPL_Gz: MdPL Gaussian-zipped artifacts - MdPL_Surrogate: MdPL surrogate models and logs - Sensitivity_Analysis: Parameter sensitivity results - Trainset: Training datasets for reproducibility Each archive is a tar.gz file. See README.md for details on unpacking and usage.

Files

README.md

Files (6.3 GB)

Name Size
md5:66989adb33aa279e593e2f8b533c7054
276.7 MB Download
md5:92cf531e54a487c045f37660502cfe53
78.1 MB Download
md5:42976c7f6163db03b9a33924bdc752f0
7.9 MB Download
md5:c1224ba85c88cdc9fd951a6a0e59efa6
23.7 MB Download
md5:35817325fb8aed830991a86f4369a908
47.4 MB Download
md5:3242a5f17116c08f46c5cf54c9e19a13
51.4 MB Download
md5:6ee7d6fa99c9db4aef066f0d091124b0
4.4 GB Download
md5:8e7eb3445ab673646c3a328e630bb3eb
533.3 MB Download
md5:8ac227ea819e7bc373501e1b96a16d10
1.4 kB Preview Download
md5:330dea45289abd0c837a0c753fcf9650
93.4 MB Download
md5:e74f45d1cca1ad016881b6f6a0a462a1
720.3 MB Download