Recollection versus Imagination: Exploring Human Memory and Cognition via Neural Language Models

@inproceedings{Sap2020RecollectionVI,
  title={Recollection versus Imagination: Exploring Human Memory and Cognition via Neural Language Models},
  author={Maarten Sap and Eric Horvitz and Yejin Choi and Noah A. Smith and James W. Pennebaker},
  booktitle={ACL},
  year={2020}
}
We investigate the use of NLP as a measure of the cognitive processes involved in storytelling, contrasting imagination and recollection of events. To facilitate this, we collect and release Hippocorpus, a dataset of 7,000 stories about imagined and recalled events. We introduce a measure of narrative flow and use this to examine the narratives for imagined and recalled events. Additionally, we measure the differential recruitment of knowledge attributed to semantic memory versus episodic… 
View on ACL
homes.cs.washington.edu
9 Citations
Background Citations
5
Methods Citations
5

Figures and Tables from this paper

9 Citations

Computational Lens on Cognition: Study Of Autobiographical Versus Imagined Stories With Large-Scale Language Models
TLDR
The study of differences in the narrative flow of events in autobiographical versus imagined stories using GPT-3, one of the largest neural language models created to date, finds that imagined stories have higher sequentiality than autobiography, and that the sequentiality of autobiographical stories is higher when they are retold than when freshly recalled.
Event segmentation reveals working memory forgetting rate
Memory and Knowledge Augmented Language Models for Inferring Salience in Long-Form Stories
TLDR
This paper improves the standard transformer language model by incorporating an external knowledgebase and adding a memory mechanism to enhance performance on longer works and uses a novel approach to derive salience annotation using chapter-aligned summaries from the Shmoop corpus for classic literary works.
A Neural Conversation Generation Model via Equivalent Shared Memory Investigation
TLDR
A novel reading and memory framework called Deep Reading Memory Network (DRMN) which is capable of remembering useful information of similar conversations for improving utterance generation is proposed and applied to two large-scale conversation datasets of justice and e-commerce fields.
Can We Map Culture?
Images that convert culture into physical space have a durable appeal, and numbers make it possible to literalize a spatial representation of culture by measuring the “distances” between cultural…
Uncovering Surprising Event Boundaries in Narratives
TLDR
The FED metric (fine-grained evaluation of dialog), an automatic evaluation metric which uses DialoGPT, without any fine-tuning or supervision is introduced, which attains moderate to strong correlation with human judgement at both levels.
Narrative Datasets through the Lenses of NLP and HCI
TLDR
This short paper compares existing value systems and approaches in NLP and HCI for collecting narrative data, and points towards contextuality, positionality, dataset size, and open research design as central points of difference and windows for collaboration when studying narratives.
MVP: Multi-task Supervised Pre-training for Natural Language Generation
TLDR
This work proposes M ulti-task super V ised P re-training ( MVP) for natural language generation, and collects a labeled pre-training corpus from 45 datasets over seven generation tasks to pre-train the text generation model MVP.
Broadening Access to Text Analysis by Describing Uncertainty: Preliminary White Paper
—We describe work in progress on a project that aims to quantify the distortion introduced into real research questions by two kinds of noise: 1) OCR error and 2) inclusion of paratext that isn’t…

References

SHOWING 1-10 OF 44 REFERENCES
Is memory for remembering? Recollection as a form of episodic hypothetical thinking
TLDR
It is claimed that remembering is a particular operation of a cognitive system that permits the flexible recombination of different components of encoded traces into representations of possible past events that might or might not have occurred, in the service of constructing mental simulations of possible future events.
Neurophysiological correlates of memory for experienced and imagined events
Misinformation Effects in Recall: Creating False Memories through Repeated Retrieval
Abstract In two experiments subjects viewed slides depicting a crime and then received a narrative containing misleading information about some items in the slides. Recall instructions were…
Semantic memory
Our concepts about objects, states, and events are stored in a cognitive structure termed semantic memory. There are several types of neurologic disorders that may cause impairments of semantic…
How schema and novelty augment memory formation
The Handbook of Emotion and Memory : Research and Theory
TLDR
This book discusses the role of affect and emotion in the formation of memory and the role that emotion plays in the development of memory impairment in patients suffering from Multiple Personality Disorder.
Literary Event Detection
TLDR
This work presents a new dataset of literary events—events that are depicted as taking place within the imagined space of a novel and applies this model to a corpus of novels split across two dimensions—prestige and popularity—and demonstrates that there are statistically significant differences in the distribution of events for prestige.
ATOMIC: An Atlas of Machine Commonsense for If-Then Reasoning
TLDR
Experimental results demonstrate that multitask models that incorporate the hierarchical structure of if-then relation types lead to more accurate inference compared to models trained in isolation, as measured by both automatic and human evaluation.
Remembering: A Study in Experimental and Social Psychology
Part I. Experimental Studies: 2. Experiment in psychology 3. Experiments on perceiving III Experiments on imaging 4-8. Experiments on remembering: (a) The method of description (b) The method of…
A Probabilistic Earley Parser as a Psycholinguistic Model
TLDR
Under grammatical assumptions supported by corpus-frequency data, the operation of Stolcke's probabilistic Earley parser correctly predicts processing phenomena associated with garden path structural ambiguity and with the subject/object relative asymmetry.
...
...