Evaluating LLM Understanding via Structured Tabular Decision Simulations
This addresses the need for better evaluation protocols for LLM understanding, which is crucial for AI researchers and developers, though it is incremental in proposing a new framework rather than a breakthrough.
The paper tackled the problem of evaluating whether large language models (LLMs) have genuine understanding beyond predictive accuracy by introducing Structured Tabular Decision Simulations (STaDS), a suite of decision settings, and found that most models struggle with consistent accuracy across domains and show mismatches between rationales and predictions.
Large language models (LLMs) often achieve impressive predictive accuracy, yet correctness alone does not imply genuine understanding. True LLM understanding, analogous to human expertise, requires making consistent, well-founded decisions across multiple instances and diverse domains, relying on relevant and domain-grounded decision factors. We introduce Structured Tabular Decision Simulations (STaDS), a suite of expert-like decision settings that evaluate LLMs as if they were professionals undertaking structured decision ``exams''. In this context, understanding is defined as the ability to identify and rely on the correct decision factors, features that determine outcomes within a domain. STaDS jointly assesses understanding through: (i) question and instruction comprehension, (ii) knowledge-based prediction, and (iii) reliance on relevant decision factors. By analyzing 9 frontier LLMs across 15 diverse decision settings, we find that (a) most models struggle to achieve consistently strong accuracy across diverse domains; (b) models can be accurate yet globally unfaithful, and there are frequent mismatches between stated rationales and factors driving predictions. Our findings highlight the need for global-level understanding evaluation protocols and advocate for novel frameworks that go beyond accuracy to enhance LLMs' understanding ability.