From the source
Lead story
Ask your AI
Top stories
Models & availability
Latest
LLMs Are Not (Consistently) Bayesian: Quantifying Internal (In)consistencies of LLMs’ Probabilistic Beliefs
Study finds LLMs' probabilistic belief updates are internally inconsistent.
Apple researchers introduce a technique to study LLMs as information processing rules, using the deviation from Bayesian updates to measure internal inconsistencies in how LLMs update probabilistic beliefs from evidence.
Their experiments show that non-Bayesian heuristic updates often outperform exact Bayesian updates in downstream tasks, suggesting LLMs' probabilistic world models are misspecified.
The measure can provide diagnostics for issues in LLM-powered inferential systems.
From the source
We introduce the novel technique of studying LLMs as information processing rules and utilize the information processing gap—the deviation from Bayes updates—to study the internal (in)consistencies of how LLMs update their probabilistic beliefs from evidence.
machinelearning.apple.com
