# Are Pretrained Language Models Symbolic Reasoners Over Knowledge?

arXiv: https://arxiv.org/abs/2006.10413 · HTML: https://mlsanity.com/paper/2006.10413/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 18 Jun 2020.

## Built on by (most cited first, 3 of 3)

- [Pushing the Limits of Rule Reasoning in Transformers through Natural Language Satisfiability](https://mlsanity.com/paper/2112.09054/) (2021)
- [Evaluating Step-by-Step Reasoning through Symbolic Verification](https://mlsanity.com/paper/2212.08686/) (2022)
- [Planning with Logical Graph-based Language Model for Instruction Generation](https://mlsanity.com/paper/2308.13782/) (2023)

## Abstract (arXiv)

How can pretrained language models (PLMs) learn factual knowledge from the training set? We investigate the two most important mechanisms: reasoning and memorization. Prior work has attempted to quantify the number of facts PLMs learn, but we present, using synthetic data, the first study that investigates the causal relation between facts present in training and facts learned by the PLM. For reasoning, we show that PLMs seem to learn to apply some symbolic reasoning rules correctly but struggle with others, including two-hop reasoning. Further analysis suggests that even the application of learned reasoning rules is flawed. For memorization, we identify schema conformity (facts systematically supported by other facts) and frequency as key factors for its success.