Home

TensorLog is a probabilistic first-order logic that has been integrated with neural networks. In TensorLog, queries are compiled into differentiable functions in a neural-network infrastructure such as Tensorflow or Theano. This leads to a close integration of probabilistic logical reasoning with deep-learning infrastructure: in particular, it enables high-performance deep learning frameworks to be used for tuning the parameters of a probabilistic logic. TensorLog scales to problems involving hundreds of thousands of knowledge-base triples and tens of thousands of examples.

TensorLog is descended from ProPPR and uses many of the same ideas.

Further reading:

• A paper on TensorLog.
• About the TensorLog Database
• About the TensorLog Program
• About TensorLog Dataset
• Using the TensorLog Interpreter to run queries interactively from Python.
• Running a simple Experiment using TensorLog's native learner.
• Using TensorLog with TensorFlow: Getting Started
• Using TensorLog with TensorFlow: More Tricks
  • to add: proof_count,builder.schema, plugins, inputs