aiqbench analysis
This research evaluates whether large language models inherently prefer simpler explanations (Occam's Razor) during inductive and abductive reasoning tasks. The study introduces a novel framework and automated metric to quantify parsimony, revealing that while SOTA models handle simple scenarios well, they struggle with complex world models even when enhanced with techniques like RLVR.
Do Language Models Follow Occam's Razor? An Evaluation of Parsimony in Inductive and Abductive Reasoning
arXiv:2509.03345v3 Announce Type: replace Abstract: Non-deductive reasoning, encompassing inductive and abductive reasoning, is essential in addressing complex real-world questions. One key feature of inductive and abductive reasoning is that there are many valid hypotheses; the simplest ones (those that adhere to Occam's Razor) are often most useful. However, this aspect is ignored in recent work that evaluates the non-deductive reasoning capabilities of large language models (LLMs). This work fills this gap, focusing on understanding whether the inductive and abductive reasoning capabilities of LLMs adhere to Occam's Razor, while also examining the correctness of their reasoning. To accomplish this goal, we introduce a framework to synthetically generate reasoning questions that (a) require inductive reasoning and abductive reasoning simultaneously; (b) is readily extended to produce any abductive/inductive reasoning question expressible in first-order logic. The task for the intelligent agent is to produce hypotheses to explain observations under a given world model. We also propose a new automated metric to assess whether hypotheses quantitatively adhere to Occam's Razor; those hypotheses that are correct and simplest are considered high-quality. Our findings on state-of-the-art LLMs suggest that LLMs can perform inductive and abductive reasoning in simple scenarios, but struggle with complex world models and with producing high-quality hypotheses, even with popular reasoning-enhancing techniques such as in-context learning and RLVR.
Title:Do Language Models Follow Occam's Razor? An Evaluation of Parsimony in Inductive and Abductive Reasoning
Abstract:Non-deductive reasoning, encompassing inductive and abductive reasoning, is essential in addressing complex real-world questions. One key feature of inductive and abductive reasoning is that there are many valid hypotheses; the simplest ones (those that adhere to Occam's Razor) are often most useful. However, this aspect is ignored in recent work that evaluates the non-deductive reasoning capabilities of large language models (LLMs). This work fills this gap, focusing on understanding whether the inductive and abductive reasoning capabilities of LLMs adhere to Occam's Razor, while also examining the correctness of their reasoning. To accomplish this goal, we introduce a framework to synthetically generate reasoning questions that (a) require inductive reasoning and abductive reasoning simultaneously; (b) is readily extended to produce any abductive/inductive reasoning question expressible in first-order logic. The task for the intelligent agent is to produce hypotheses to explain observations under a given world model. We also propose a new automated metric to assess whether hypotheses quantitatively adhere to Occam's Razor; those hypotheses that are correct and simplest are considered high-quality. Our findings on state-of-the-art LLMs suggest that LLMs can perform inductive and abductive reasoning in simple scenarios, but struggle with complex world models and with producing high-quality hypotheses, even with popular reasoning-enhancing techniques such as in-context learning and RLVR.
Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
arXiv:2509.03345 [cs.AI]
(or arXiv:2509.03345v3 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2509.03345 Focus to learn more arXiv-issued DOI via DataCite
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.
Full text under the publisher's open licence · 320 words
Related coverage
Comment on Measuring Reward-Seeking by Instilling Contrastive Beliefs paper from mechanistic interpretability perspective
LessWrong · Jul 22, 2026
Deconfusing Corrigibility
AI Alignment Forum · Oct 17, 2026
Angles of attack for continual learning safety
LessWrong · Jun 16, 2026
Latent reasoning architectures would likely undermine CoT, our strongest oversight tool
AI Alignment Forum · Sep 23, 2026