Improving Neural Logic Machines via Failure Reflection
Nanyang Technological University · Tianjin University, Nanyang Technical University · National University of Singapore · Hong Kong Polytechnic University · Continental Automotive Singapore · School of Computer Science and Engineering, Nanyang Technological University · Nanyang Technology University, Singapore
PDF 由论文原始站点提供，PaperCompass 不保存论文文件。
摘要
Reasoning is a fundamental ability towards artificial general intelligence (AGI). Fueled by the success of deep learning, the neural logic machines models (NLMs) have introduced novel neural-symbolic structures and demonstrate great performance and generalization on reasoning and decision-making tasks. However, the original training approaches of the NLMs are still far from perfect, the models would repeat similar mistakes during the training process which leads to sub-optimal performance. To mitigate this issue, we present a novel framework named Failure Reflection Guided Regularizer (FRGR). FRGR first dynamically identifies and summarizes the root cause if the model repeats similar mistakes during training. Then it penalizes the model if it makes similar mistakes in future training iterations. In this way, the model is expected to avoid repeating errors of similar root causes and converge faster to a better-performed optimum. Experimental results on multiple relational reasoning and decision-making tasks demonstrate the effectiveness of FRGR in improving performance, generalization, training efficiency, and data efficiency.