TRAC: A Textual Benchmark for Reasoning about Actions and Change
This addresses the problem of assessing structural generalization in AI for dynamic environments, but it is incremental as it builds on existing benchmarks and methods.
The authors introduced TRAC, a textual benchmark for evaluating language models on reasoning about actions and change, and found that current high-performing transformers struggle with the tasks, indicating a need for further improvements.
Reasoning about actions and change (RAC) is essential to understand and interact with the ever-changing environment. Previous AI research has shown the importance of fundamental and indispensable knowledge of actions, i.e., preconditions and effects. However, traditional methods rely on logical formalization which hinders practical applications. With recent transformer-based language models (LMs), reasoning over text is desirable and seemingly feasible, leading to the question of whether LMs can effectively and efficiently learn to solve RAC problems. We propose four essential RAC tasks as a comprehensive textual benchmark and generate problems in a way that minimizes the influence of other linguistic requirements (e.g., grounding) to focus on RAC. The resulting benchmark, TRAC, encompassing problems of various complexities, facilitates a more granular evaluation of LMs, precisely targeting the structural generalization ability much needed for RAC. Experiments with three high-performing transformers indicates that additional efforts are needed to tackle challenges raised by TRAC.