Lgmt tool finds reasoning flaws in large language models
LGMT: A Tool for Testing the Logical Reasoning Reliability of LLMs
Artificial Intelligence
Summary
Many large language models can give answers that contradict each other when the same question is asked in different ways. The authors created LGMT, a tool that tests these models for logical consistency without needing correct answers beforehand. It does this by making versions of a question that should mean the same thing and checking if the model answers them consistently. They tested six popular models and found that about 20% to 30% of the time, they gave inconsistent answers.
What this means in practice
- •For software developers: Check language models used in software for logical consistency to reduce contradictory outputs.
- •For ai quality assurance teams: Verify LLMs across logically equivalent inputs automatically to find and fix reasoning flaws before deployment.
Authors
Zenghui Zhou, Xiaoke Fang, Zheng Zheng
Abstract
Large Language Models (LLMs) increasingly serve as reasoning components in modern software systems, yet a model that answers a logical query correctly can contradict itself when the same problem is restated in an equivalent form. Static benchmarks evaluate queries in isolation and therefore fail to expose such inconsistencies. To address this limitation, we present LGMT, an open-source tool that detects reasoning inconsistencies through metamorphic relations grounded in first-order logic. Given a source test case in premise-conclusion form, LGMT generates logically equivalent follow-up cases and checks whether the model's entailment predictions agree across the pair, without ground-truth labels. The tool provides 20 built-in metamorphic relations, supports custom ones verified by the Z3 theorem prover, and runs from a CLI or a web studio, including an offline mock mode requiring no API key. Across six widely used LLMs on a bundled benchmark of 3,091 metamorphic groups, LGMT reveals inconsistency rates ranging from 20.84% to 29.44%. LGMT is publicly available on GitHub at https://github.com/MaxwelsDonc/lgmt-tool, with a demonstration video at https://youtu.be/gou4opptu1U.