Kullback-Leibler Divergence
Version 1.0.0 · Updated 2026-07-30
CORE DEFINITION
A metric used to measure the difference between two probability distributions (e.g., 'your beliefs' and 'the real world'). The larger the KL divergence, the more your model deviates from reality, and the more 'information' you need to correct it. - Scaffolding role: A measure of cognitive surprise. - The essence of learning is to minimize KL divergence. When you encounter something that makes you 'extremely surprised' (KL divergence spikes), that is the best moment for cognitive upgrade. Do not ignore surprise; surprise means your internal model is severely inconsistent with the external world, which is a golden opportunity to correct cognition.
SCAFFOLDING EFFECT
Reduce cognitive load
A measure of cognitive surprise. - The essence of learning is to minimize KL divergence. When you encounter something that makes you 'extremely surprised' (KL divergence spikes), that is the best moment for cognitive upgrade. Do not ignore surprise; surprise means your internal model is severely inconsistent with the external world, which is a golden opportunity to correct cognition.
Anchor fast decisions
A metric for measuring the difference between two probability distributions (relative entropy). The larger the KL divergence, the more the model deviates from reality, and the more information is needed to correct it. The mechanism can be extended as a measure of cognitive surprise—the essence of learning is to minimize KL divergence; when encountering something extremely surprising (KL spikes), it indicates a severe mismatch between the internal model and the external world, presenting a golden opportunity to correct cognition.
MINIMUM ACTION
In progress 0/4Practice this model in one real situation:
account_treeGenealogyexpand_more
menu_bookReferencesexpand_more
Source support: Explicit
- zh.wikipedia.orghttps://zh.wikipedia.org/wiki/%E7%9B%B8%E5%AF%B9%E7%86%B5verified
PRIVATE NOTES · Only visible to you
SAVED Q&A
ENTRY Q&A · Private saving available
Ask with a clear boundary
thinkingmodels answers from published entry context only.
Your question is sent to thinkingmodels. The answer uses public entry context only.
RELATED MODELS