Toward Mechanistic Explanation of Deductive Reasoning in Language Models
This provides a mechanistic explanation for deductive reasoning in language models, which is incremental as it builds on existing capabilities to uncover internal mechanisms.
The paper tackled the problem of understanding how language models perform deductive reasoning by showing that a small model can learn underlying rules, not just statistical patterns, and found that induction heads are central to rule completion and chaining in logical inference.
Recent large language models have demonstrated relevant capabilities in solving problems that require logical reasoning; however, the corresponding internal mechanisms remain largely unexplored. In this paper, we show that a small language model can solve a deductive reasoning task by learning the underlying rules (rather than operating as a statistical learner). A low-level explanation of its internal representations and computational circuits is then provided. Our findings reveal that induction heads play a central role in the implementation of the rule completion and rule chaining steps involved in the logical inference required by the task.