gym-saturation 0.12.2

Gymnasium environments for saturation provers

gym-saturation

gym-saturation is a collection of Gymnasium environments for reinforcement learning (RL) agents guiding saturation-style automated theorem provers (ATPs) based on the given clause algorithm.

There are two environments in gym-saturation following the same API: SaturationEnv: VampireEnv — for Vampire prover, and IProverEnv — for iProver.

gym-saturation can be interesting for RL practitioners willing to apply their experience to theorem proving without coding all the logic-related stuff themselves.

In particular, ATPs serving as gym-saturation backends incapsulate parsing the input formal language (usually, one of the TPTP (Thousands of Problems for Theorem Provers) library), transforming the input formulae to the clausal normal form, and logic inference using rules such as resolution and superposition.

How to Install

Attention!

If you want to use VampireEnv you should have a Vampire binary on your machine. For example, download the latest release.

To use IProverEnv, please download a stable iProver release or build it from this commit.

The best way to install this package is to use pip:

pip install gym-saturation

Another option is to use conda:

conda install -c conda-forge gym-saturation

One can also run it in a Docker container (pre-packed with vampire and iproveropt binaries):

docker build -t gym-saturation https://github.com/inpefess/gym-saturation.git
docker run --rm -p 8888:8888 --name gym-saturation -d gym-saturation

And then naviagate to http://localhost:8888/lab/tree/example.ipynb in your browser.

How to use

One can use gym-saturation environments as any other Gymnasium environment:

import gym_saturation
import gymnasium

env = gymnasium.make("Vampire-v0")  # or "iProver-v0"
# skip this line to use the default problem
env.set_task("a-TPTP-problem-filename")
observation, info = env.reset()
terminated, truncated = False, False
while not (terminated or truncated):
    # apply policy (a random action here)
    action = env.action_space.sample()
    observation, reward, terminated, truncated, info = env.step(action)
env.close()

Or have a look at the basic tutorial.

For a bit more comprehensive experiments, please see this project.

More Documentation

More documentation can be found here.

Related Projects

gym-saturation is compatible with RL-frameworks such as Ray RLlib and can leverage code embeddings such as CodeBERT.

Other projects using RL-guidance for ATPs include:

• TRAIL

• FLoP (see the paper for more details)

• lazyCoP (see the paper for more details)

Other projects not using RL per se, but iterating a supervised learning procedure instead:

• ENIGMA (several repos, e.g. this one for iProver; see the paper for others)

• Deepire

How to Contribute

Please follow the contribution guide while adhering to the code of conduct.

How to Cite

If you are writing a research paper and want to cite gym-saturation, please use the following DOI.