Releases: sammsaski/VVN
Release list
FormaliSE 2025
State of the code (before all clean up for artifact evaluation, etc.) for submission to FormaliSE 2025.
Keeping this release as a safeguard in case things go wrong when cleaning up repo for AE.
main branch pre-FormaliSE submission
Self-explanatory.
SEFM 2024 v0.2.0
Added code and generated plots that illustrate the computed output ranges.
Still to do :
- Run verification for some randomly selected samples on UCF11 model
- Check the range plots for verified samples
- Retrain + rerun verification on ZoomOut8f
- Cleanup repository
- Update README with reproducibility instructions
SEFM 2024 v0.1.0
Added code and generated plots that illustrate the computed output ranges.
Still to do :
- Finish the feature map experiment
- Run exact for ZoomIn + ZoomOut with frames=[4, 8]
- Run verification for some randomly selected samples on UCF11 model
- Check the range plots for verified samples
- Retrain + rerun verification on ZoomOut8f
- Cleanup repository
- Update README with reproducibility instructions
SEFM 2024 v0.0.2
Removed unnecessary files from repository that were in previous release.
Still to do :
- Finish the feature map experiment
- Run exact for ZoomIn + ZoomOut with frames=[4, 8]
- Run verification for some randomly selected samples on UCF11 model
- Check the range plots for verified samples
- Retrain + rerun verification on ZoomOut8f
- Cleanup repository
- Update README with reproducibility instructions
SEFM 2024 v0.0.1
Finished running verification on ZoomIn/ZoomOut datasets with frames=[4, 8, 16]. with relax and with frames[4, 8] with approx.
Still to do :
- Finish the feature map experiment
- Run exact for ZoomIn + ZoomOut with frames=[4, 8]
- Run verification for some randomly selected samples on UCF11 model
- Check the range plots for verified samples
- Retrain + rerun verification on ZoomOut8f
- Cleanup repository
- Update README with reproducibility instructions
Inorder Experiments
This was the state of the codebase when I finally finished up running the inorder experiments.
Pre-cleanup
Repository was getting too messy with different datasets, model types, etc. I don't want to lose all previous work so I'm tagging it as a safety net given that the repository is purely for research purposes. Please ignore.
CS6315Final
State of the repository as of submission for the CS6315 final project.