Rationality Check! Benchmarking the Rationality of Large Language Models

2025/09/18 by Zhou, Zhilun, Jing Yi Wang, Wang, Jing Yi +10
Computer Science · Medicine · #Artificial Intelligence (cs.AI) #Artificial Intelligence in Healthcare and Education #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2509.14546

openalex publication_date 2025/09/18 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Large language models (LLMs), a recent advance in deep learning and machine intelligence, have manifested astonishing capacities, now considered among the most promising for artificial general intelligence. With human-like capabilities, LLMs have been used to simulate humans and serve as AI assistants across many applications. As a result, great concern has arisen about whether and under what circumstances LLMs think and behave like real human agents. Rationality is among the most important concepts in assessing human behavior, both in thinking (i.e., theoretical rationality) and in taking action (i.e., practical rationality). In this work, we propose the first benchmark for evaluating the omnibus rationality of LLMs, covering a wide range of domains and LLMs. The benchmark includes an easy-to-use toolkit, extensive experimental results, and analysis that illuminates where LLMs converge and diverge from idealized human rationality. We believe the benchmark can serve as a foundational tool for both developers and users of LLMs.

Citations

• DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model
• Large Language Models Empowered Agent-based Modeling and Simulation: A Survey and Perspectives
• Exploring Collaboration Mechanisms for LLM Agents: A Social Psychology View
• The Rise and Potential of Large Language Model Based Agents: A Survey
• Trustworthy LLMs: a Survey and Guideline for Evaluating Large Language Models' Alignment
• Ghost in the Minecraft: Generally Capable Agents for Open-World Environments via Large Language Models with Text-based Knowledge and Memory
• The Emergence of Economic Rationality of GPT
• LLaMA: Open and Efficient Foundation Language Models
• Large Language Models Fail on Trivial Alterations to Theory-of-Mind Tasks
• A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivity
• Out of One, Many: Using Language Models to Simulate Human Samples
• Using Large Language Models to Simulate Multiple Humans and Replicate Human Subject Studies
• Training language models to follow instructions with human feedback
• Measuring Mathematical Problem Solving With the MATH Dataset
• Critical Thinking for Language Models
• Measuring Massive Multitask Language Understanding
• Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being.
• Decision-Making Style: The Development and Assessment of a New Measure
• The Efficient Assessment of Need for Cognition
• MLAgentBench: Evaluating Language Agents on Machine Learning Experimentation
• EconAgent: Large Language Model-Empowered Agents for Simulating Macroeconomic Activities
• Development and Validation of the Scientific Reasoning Scale

Related

• Comparing Rationality Between Large Language Models and Humans: Insights and Open Questions
• (Ir)rationality and Cognitive Biases in Large Language Models
• Are Machine Rationales (Not) Useful to Humans? Measuring and Improving Human Utility of Free-Text Rationales
• Advances in LLMs with Focus on Reasoning, Adaptability, Efficiency and Ethics
• Domain Specific Benchmarks for Evaluating Multimodal Large Language Models