On the Tip of the Tongue: Analyzing Conceptual Representation in Large Language Models with Reverse-Dictionary Probe
This work addresses the challenge of understanding and enhancing reasoning in LLMs for AI researchers, though it is incremental as it applies an existing method to a new task.
The study tackled the problem of probing large language models' conceptual reasoning by using a reverse-dictionary task, where models achieved high accuracy in generating terms from descriptions, and this ability predicted performance on broader reasoning benchmarks.
Probing and enhancing large language models' reasoning capacity remains a crucial open question. Here we re-purpose the reverse dictionary task as a case study to probe LLMs' capacity for conceptual inference. We use in-context learning to guide the models to generate the term for an object concept implied in a linguistic description. Models robustly achieve high accuracy in this task, and their representation space encodes information about object categories and fine-grained features. Further experiments suggest that the conceptual inference ability as probed by the reverse-dictionary task predicts model's general reasoning performance across multiple benchmarks, despite similar syntactic generalization behaviors across models. Explorative analyses suggest that prompting LLMs with description$\Rightarrow$word examples may induce generalization beyond surface-level differences in task construals and facilitate models on broader commonsense reasoning problems.