Metaphor and Large Language Models: When Surface Features Matter More than Deep Understanding
This work addresses the need for realistic evaluation frameworks in NLP for metaphor interpretation, highlighting limitations in LLMs' capabilities, which is incremental as it builds on prior single-dataset studies.
The paper tackled the problem of evaluating Large Language Models' (LLMs) ability to interpret metaphors by conducting experiments across multiple datasets and tasks, finding that LLMs' performance is more influenced by surface features like lexical overlap and sentence length than by metaphorical content, with results indicating no emergent understanding of figurative language.
This paper presents a comprehensive evaluation of the capabilities of Large Language Models (LLMs) in metaphor interpretation across multiple datasets, tasks, and prompt configurations. Although metaphor processing has gained significant attention in Natural Language Processing (NLP), previous research has been limited to single-dataset evaluations and specific task settings, often using artificially constructed data through lexical replacement. We address these limitations by conducting extensive experiments using diverse publicly available datasets with inference and metaphor annotations, focusing on Natural Language Inference (NLI) and Question Answering (QA) tasks. The results indicate that LLMs' performance is more influenced by features like lexical overlap and sentence length than by metaphorical content, demonstrating that any alleged emergent abilities of LLMs to understand metaphorical language are the result of a combination of surface-level features, in-context learning, and linguistic knowledge. This work provides critical insights into the current capabilities and limitations of LLMs in processing figurative language, highlighting the need for more realistic evaluation frameworks in metaphor interpretation tasks. Data and code are publicly available.