Simulating Action Dynamics with Neural Process Networks
arXiv:1711.05313
Abstract
Understanding procedural language requires anticipating the causal effects of actions, even when they are not explicitly stated. In this work, we introduce Neural Process Networks to understand procedural text through (neural) simulation of action dynamics. Our model complements existing memory architectures with dynamic entity tracking by explicitly modeling actions as state transformers. The model updates the states of the entities by executing learned action operators. Empirical results demonstrate that our proposed model can reason about the unstated causal effects of actions, allowing it to provide more accurate contextual information for understanding and generating procedural text, all while offering more interpretable internal representations than existing alternatives.
References in corpus (10)
- Efficient Estimation of Word Representations in Vector Space
- Distributed Representations of Words and Phrases and their Compositionality
- Sequence to Sequence Learning with Neural Networks
- End-To-End Memory Networks
- The Goldilocks Principle: Reading Children's Books with Explicit Memory Representations
- Hadamard Product for Low-rank Bilinear Pooling
- Tracking the World State with Recurrent Entity Networks
- Recurrent Environment Simulators
- Query-Reduction Networks for Question Answering
- Learning a Natural Language Interface with Neural Programmer
Cited by in corpus (11)
- On the Opportunities and Risks of Foundation Models
- What Happened? Leveraging VerbNet to Predict the Effects of Actions in Procedural Text
- Tracking State Changes in Procedural Text: A Challenge Dataset and Models for Process Paragraph Comprehension
- RecipeQA: A Challenge Dataset for Multimodal Comprehension of Cooking Recipes
- Reasoning about Actions and State Changes by Injecting Commonsense Knowledge
- Discourse-Aware Neural Rewards for Coherent Text Generation
- Modeling Label Semantics for Predicting Emotional Reactions
- Tracking entities in technical procedures -- a new dataset and baselines
- Designing Multimodal Datasets for NLP Challenges
- Action Recognition and State Change Prediction in a Recipe Understanding Task Using a Lightweight Neural Network Model
- SHARE: a System for Hierarchical Assistive Recipe Editing