Bayesian teaching enables probabilistic reasoning in large language models

doi:10.1038/s41467-025-67998-6

Bayesian teaching enables probabilistic reasoning in large language models

2026-01-07

0

OA

AI

Sjoerd van Steenkiste *

DOI：10.1038/s41467-025-67998-6

Original

Save

View PDF

Abstract

En 中文

Large language models (LLMs) are increasingly used as agents that interact with users and with the world. To do so successfully, LLMs must construct representations of the world and form probabilistic beliefs about them. To provide personalized recommendations, for example, the LLM needs to infer a user’s preferences from their behavior over multiple interactions. The Bayesian inference framework lays out the optimal way for an agent to update its beliefs as it receives new information. We first show that LLMs fall far short of the standard defined by the Bayesian framework. We then show that by teaching LLMs to mimic the predictions of the normative Bayesian model, we can dramatically improve their ability to update their beliefs; this ability generalizes to new tasks. We conclude that LLMs can effectively learn reasoning skills from examples and generalize those skills to new domains. LLMs fail to update beliefs in a Bayesian way but can be taught to do so by mimicking a normative Bayesian model. This training yields better predictions and transferable reasoning skills across tasks.

Keywords:

Bayesian inference

probabilistic reasoning

large language models

belief updating

transferable skills

Bayesian teaching enables probabilistic reasoning in large language models

Abstract

AI Summary

Journal

Organization