This repo contains the code for PAL, a probabilistic neuro-symbolic layer for algebraic constraint satisfaction. This is a simplified implementation that focuses on the spline-case.
Check out the paper: https://arxiv.org/abs/2503.19466v1
This is an example prediction of PAL on the Constrained Stanford Drone Dataset (https://github.com/april-tools/constrained-sdd). We predict a probability distribution over the future trajectory while guaranteeing constraint-satisfaction.
Leander Kurscheidt, Paolo Morettin, Roberto Sebastiani, Andrea Passerini, Antonio Vergari, A Probabilistic Neuro-symbolic Layer for Algebraic Constraint Satisfaction, arXiv:2503.19466
Just clone it and run:
./setup.shAnd you're ready to go!
We provide an example script how to train a simple MLP on the constrained SDD-dataset. A model can be trained like this:
python pal/training/train_mlp_sdd.py --epochs 10 --init_last_layer_positive --seed 1744909132This should result in a (mean) test log-likelihood of -1.9800.
The dependency was added via subtree from https://github.com/april-tools/gasp.git into pal/wmi/gasp! update via:
git subtree pull --prefix pal/wmi/gasp https://github.com/april-tools/gasp.git main --squashpush via:
git subtree push --prefix pal/wmi/gasp https://github.com/april-tools/gasp.git main