Work

Can LLMs Disambiguate Grounded Language? The Case of PP Attachment

Year: 2025

Type: conference-paper

Abstract: We study resolution of ambiguity in prepositional phrase attachment by Large Language Models in the zero-shot setting.We evaluate a strong "plausibility" baseline derived from token probabilities of descriptions encoding alternative attachments, and explore possible improvements using additional token probabilities that reflect aspects of information structure.Error analysis suggests directions for more sophisticated tools, common-sense reasoning, world knowledge, and additional context to better resolve ambiguity.

Source: International conference Recent advances in natural language processing

DOI: 10.26615/978-954-452-098-4-020

Authors: John Blackmore, Matthew Stone

Institution: Rutgers, The State University of New Jersey

Language: English

Topic: Natural Language Processing Techniques

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

Funder: National Science Foundation

Awards: ReDDDoT Phase 1: Planning Grant: Writing Education through Design-Oriented AI, NRT-FW-HTF: Socially Cognizant Robotics for a Technology Enhanced Society (SOCRATES), The Role of Agency and Interactivity in Embodied Conversational Agents to Support Teacher Training in High Poverty Schools