The Imperfective Paradox Is Not Necessarily in Large Language Models: A Benchmark Failure Before a Model Failure

Abstract:The imperfective paradox provides a useful test of compositional semantic analysis. Recent work constructs an NLI benchmark and reports that models frequently infer completed telic events from progressive descriptions, attributing this behavior to a Teleological Bias. It further argues that prompting interventions cause a Calibration Crisis. We reexamine the benchmark and conclusions and show that it is substantially affected by conceptual and evaluation mis-specifications. We identify three conceptual mis-specifications. In particular, Aspectual Reduction affects the benchmark construction, analysis, experiments, and conclusions. Under a strict NLI standard, 76% of Group A instances do not explicitly rule out culmination. In our native-speaker annotation, 38% of Group A examples and 29% of the Group C examples were judged to permit an alternative interpretation. To control these issues and lexical variation, we construct Lexically Matched Minimal Pairs. At the evaluation level, we formulate event-semantic NLI as a Multi-step Reasoning Problem and assess both intermediate semantic decisions and final predictions. Our results show that models often do not affirm culmination but nevertheless accept the corresponding simple-past hypothesis, a pattern we characterize as Sufficiency Bias. We further show that prompting interventions produce a Decision Shift among labels without reliably improving the underlying semantic understanding and reasoning. Intermediate and oracle-guided analyses identify two additional failure modes: errors in compositional aspectual classification and Surface-form Attraction toward surface-associated answers. Our experiments on Qwen-7B with suitable prompts, GPT-5.4, and Qwen-72B provide initial evidence for the context sensitivity of aspectual classification and suggest that these models can achieve performance comparable to that of human annotators.

Computation and Language (cs.CL); Machine Learning (cs.LG)

arXiv:2608.25005 [cs.CL]

(or arXiv:2608.25005v1 [cs.CL] for this version)

https://doi.org/10.48550/arXiv.2608.25005 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.