Bayesian Teaching Enables Probabilistic Reasoning in Large Language Models
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911377350721536 |
|---|---|
| author | Qiu, Linlu Sha, Fei Allen, Kelsey Kim, Yoon Linzen, Tal van Steenkiste, Sjoerd |
| author_facet | Qiu, Linlu Sha, Fei Allen, Kelsey Kim, Yoon Linzen, Tal van Steenkiste, Sjoerd |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_17523 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Bayesian Teaching Enables Probabilistic Reasoning in Large Language Models Qiu, Linlu Sha, Fei Allen, Kelsey Kim, Yoon Linzen, Tal van Steenkiste, Sjoerd Computation and Language Artificial Intelligence 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. |
| title | Bayesian Teaching Enables Probabilistic Reasoning in Large Language Models |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2503.17523 |