# Characterizing Verbatim Short-Term Memory in Neural Language Models

arXiv: https://arxiv.org/abs/2210.13569 · HTML: https://mlsanity.com/paper/2210.13569/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 28 of its 42 references that are in the map.
- First posted on arXiv 24 Oct 2022.

## Builds on (most cited first, 28 of 28)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [Neural Machine Translation by Jointly Learning to Align and Translate](https://mlsanity.com/paper/1409.0473/) (2014)
- [Pointer Sentinel Mixture Models](https://mlsanity.com/paper/1609.07843/) (2016)
- [The Curious Case of Neural Text Degeneration](https://mlsanity.com/paper/1904.09751/) (2019)
- [In-context Learning and Induction Heads](https://mlsanity.com/paper/2209.11895/) (2022)
- [Efficient Transformers: A Survey](https://mlsanity.com/paper/2009.06732/) (2020)
- [Linear Transformers Are Secretly Fast Weight Programmers](https://mlsanity.com/paper/2102.11174/) (2021)
- [Memory Networks](https://mlsanity.com/paper/1410.3916/) (2014)
- [Assessing the Ability of LSTMs to Learn Syntax-Sensitive Dependencies](https://mlsanity.com/paper/1611.01368/) (2016)
- [Targeted Syntactic Evaluation of Language Models](https://mlsanity.com/paper/1808.09031/) (2018)
- [Incorporating Copying Mechanism in Sequence-to-Sequence Learning](https://mlsanity.com/paper/1603.06393/) (2016)
- [Colorless green recurrent networks dream hierarchically](https://mlsanity.com/paper/1803.11138/) (2018)
- [Regularizing and Optimizing LSTM Language Models](https://mlsanity.com/paper/1708.02182/) (2017)
- [Using Fast Weights to Attend to the Recent Past](https://mlsanity.com/paper/1610.06258/) (2016)
- [Neural Language Models as Psycholinguistic Subjects: Representations of Syntactic State](https://mlsanity.com/paper/1903.03260/) (2019)
- [Sharp Nearby, Fuzzy Far Away: How Neural Language Models Use Context](https://mlsanity.com/paper/1805.04623/) (2018)
- [Syntactic Structure from Deep Learning](https://mlsanity.com/paper/2004.10827/) (2020)
- [The emergence of number and syntax units in LSTM language models](https://mlsanity.com/paper/1903.07435/) (2019)
- [Learning to Transduce with Unbounded Memory](https://mlsanity.com/paper/1506.02516/) (2015)
- [What Context Features Can Transformer Language Models Use?](https://mlsanity.com/paper/2106.08367/) (2021)
- [Cognitive Psychology for Deep Neural Networks: A Shape Bias Case Study](https://mlsanity.com/paper/1706.08606/) (2017)
- [Mechanisms for Handling Nested Dependencies in Neural-Network Language Models and Humans](https://mlsanity.com/paper/2006.11098/) (2020)
- [RNNs as psycholinguistic subjects: Syntactic state and grammatical dependency](https://mlsanity.com/paper/1809.01329/) (2018)
- [Multi-scale Transformer Language Models](https://mlsanity.com/paper/2005.00581/) (2020)

## Built on by (most cited first, 8 of 8)

- [Language Model Behavior: A Comprehensive Survey](https://mlsanity.com/paper/2303.11504/) (2023)
- [Between Circuits and Chomsky: Pre-pretraining on Formal Languages Imparts Linguistic Biases](https://mlsanity.com/paper/2502.19249/) (2025)
- [Transformer Working Memory Enables Regular Language Reasoning and Natural Language Length Extrapolation](https://mlsanity.com/paper/2305.03796/) (2023)
- [Humans and language models diverge when predicting repeating text](https://mlsanity.com/paper/2310.06408/) (2023)
- [To model human linguistic prediction, make LLMs less superhuman](https://mlsanity.com/paper/2510.05141/) (2025)
- [Human-like fleeting memory improves language learning but impairs reading time prediction in transformer language models](https://mlsanity.com/paper/2508.05803/) (2025)
- [Simulating Human Memory with Language Models](https://mlsanity.com/paper/2605.25680/) (2026)
- [Synthetic Pre-pretraining Survives Scale, but Not as a Grammatical Prior](https://mlsanity.com/paper/2609.39827/) (2026)

## Abstract (arXiv)

When a language model is trained to predict natural language sequences, its prediction at each moment depends on a representation of prior context. What kind of information about the prior context can language models retrieve? We tested whether language models could retrieve the exact words that occurred previously in a text. In our paradigm, language models (transformers and an LSTM) processed English text in which a list of nouns occurred twice. We operationalized retrieval as the reduction in surprisal from the first to the second list. We found that the transformers retrieved both the identity and ordering of nouns from the first list. Further, the transformers' retrieval was markedly enhanced when they were trained on a larger corpus and with greater model depth. Lastly, their ability to index prior tokens was dependent on learned attention patterns. In contrast, the LSTM exhibited less precise retrieval, which was limited to list-initial tokens and to short intervening texts. The LSTM's retrieval was not sensitive to the order of nouns and it improved when the list was semantically coherent. We conclude that transformers implemented something akin to a working memory system that could flexibly retrieve individual token representations across arbitrary delays; conversely, the LSTM maintained a coarser and more rapidly-decaying semantic gist of prior tokens, weighted toward the earliest items.