CONDESION-BENCH: Conditional Decision-Making of Large Language Models in Compositional Action Space

TL;DR

CONDESION-BENCH is a new benchmark for evaluating large language models' ability to make decisions within complex, condition-restricted, compositional action spaces, addressing limitations of previous simplified benchmarks.

Contribution

It introduces a novel benchmark that incorporates explicit conditions and compositional actions, providing a more realistic evaluation of LLM decision-making capabilities.

Findings

Benchmark reveals LLMs' strengths and weaknesses in conditional decision-making.

Oracle-based evaluation offers rigorous assessment of decision quality and condition adherence.

Addresses limitations of previous decision-making benchmarks.

Abstract

Large language models have been widely explored as decision-support tools in high-stakes domains due to their contextual understanding and reasoning capabilities. However, existing decision-making benchmarks rely on two simplifying assumptions: actions are selected from a finite set of pre-defined candidates, and explicit conditions restricting action feasibility are not incorporated into the decision-making process. These assumptions fail to capture the compositional structure of real-world actions and the explicit conditions that constrain their validity. To address these limitations, we introduce CONDESION-BENCH, a benchmark designed to evaluate conditional decision-making in compositional action space. In CONDESION-BENCH, actions are defined as allocations to decision variables and are restricted by explicit conditions at the variable, contextual, and allocation levels. By employing oracle-based evaluation of both decision quality and condition adherence, we provide a more rigorous assessment of LLMs as decision-support tools.