Neural Program Modeling Learning Program-Specific Models that Predict Execution Features from Input Features and vice versa

Short summary

Neural program modeling learns reversible, program-specific models that predict execution features (like traces or coverage) from inputs, and vice versa, achieving up to 98.59% similarity for unseen inputs and 100% accuracy in predicting inputs for specific behaviors.

AI-generated from the title and abstract; the full text is not read.