# Reversal Curse

> The reversal curse is the failure of language models trained on a fact stated as "A is B" to answer the reversed question "what is B?", for example knowing who a celebrity's mother is but not who that woman's famous son is.

- Identifier: PTL-0171
- Category: Failure Modes & Evaluation
- Canonical URL: https://protologue.com/t/reversal-curse/
- Introduced: 2023

## Description

It shows that models do not automatically store facts in both directions. In practice, providing the relevant fact in the prompt fixes it, since in-context reasoning handles the reversal, which is one more argument for retrieval over relying on memorized knowledge.

## Related terms

- [Hallucination](https://protologue.com/t/hallucination/)
- [Knowledge Cutoff](https://protologue.com/t/knowledge-cutoff/)
- [Retrieval-Augmented Generation](https://protologue.com/t/retrieval-augmented-generation/)

## Sources

- Berglund et al. (2023). The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A". https://arxiv.org/abs/2309.12288

## Cite this entry

Protologue. (2026). Reversal Curse. In Protologue: A Taxonomy of Prompting and LLM Techniques (v1.4.0, PTL-0171). https://protologue.com/t/reversal-curse/

License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)