# When Models Know More Than They Say: Probing Analogical Reasoning in LLMs

arXiv: https://arxiv.org/abs/2604.03877 · HTML: https://mlsanity.com/paper/2604.03877/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning.
- Builds on 10 of its 30 references that are in the map.
- First posted on arXiv 4 Apr 2026.

## Placed in

- LLM reasoning / [Reasoning benchmarks and evaluation](https://mlsanity.com/lr/reasoning-benchmarks-and-evaluation/)

## Builds on (most cited first, 10 of 10)

- [Locating and Editing Factual Associations in GPT](https://mlsanity.com/paper/2202.05262/) (2022)
- [Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models](https://mlsanity.com/paper/2206.04615/) (2022)
- [Transformer Feed-Forward Layers Are Key-Value Memories](https://mlsanity.com/paper/2012.14913/) (2020)
- [FANToM: A Benchmark for Stress-testing Machine Theory of Mind in Interactions](https://mlsanity.com/paper/2310.15421/) (2023)
- [Large Language Models as Analogical Reasoners](https://mlsanity.com/paper/2310.01714/) (2023)
- [One Thousand and One Pairs: A "novel" challenge for long-context language models](https://mlsanity.com/paper/2406.16264/) (2024)
- [An Annotated Dataset of Coreference in English Literature](https://mlsanity.com/paper/1912.01140/) (2019)
- [Decoding Probing: Revealing Internal Linguistic Structures in Neural Language Models using Minimal Pairs](https://mlsanity.com/paper/2403.17299/) (2024)
- [Measuring Information Propagation in Literary Social Networks](https://mlsanity.com/paper/2004.13980/) (2020)
- [NarraBench: A Comprehensive Framework for Narrative Benchmarking](https://mlsanity.com/paper/2510.09869/) (2025)

## Abstract (arXiv)

Analogical reasoning is a core cognitive faculty essential for narrative understanding. While LLMs perform well when surface and structural cues align, they struggle in cases where an analogy is not apparent on the surface but requires latent information, suggesting limitations in abstraction and generalisation. In this paper we compare a model's probed representations with its prompted performance at detecting narrative analogies, revealing an asymmetry: for rhetorical analogies, probing significantly outperforms prompting in open-source models, while for narrative analogies, they achieve a similar (low) performance. This suggests that the relationship between internal representations and prompted behavior is task-dependent and may reflect limitations in how prompting accesses available information.