Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Computational Models to Study Language Processing in the Human Brain: A Survey

Shaonan Wang, Jingyuan Sun, Yunhao Zhang, Nan Lin, Marie-Francine Moens, Chengqing Zong

TL;DR

Analysis of efforts in using computational models for brain research reveals that no single model outperforms others on all datasets, underscoring the need for rich testing datasets and rigid experimental control to draw robust conclusions in studies involving computational models.

Abstract

Despite differing from the human language processing mechanism in implementation and algorithms, current language models demonstrate remarkable human-like or surpassing language capabilities. Should computational language models be employed in studying the brain, and if so, when and how? To delve into this topic, this paper reviews efforts in using computational models for brain research, highlighting emerging trends. To ensure a fair comparison, the paper evaluates various computational models using consistent metrics on the same dataset. Our analysis reveals that no single model outperforms others on all datasets, underscoring the need for rich testing datasets and rigid experimental control to draw robust conclusions in studies involving computational models.

Computational Models to Study Language Processing in the Human Brain: A Survey

TL;DR

Analysis of efforts in using computational models for brain research reveals that no single model outperforms others on all datasets, underscoring the need for rich testing datasets and rigid experimental control to draw robust conclusions in studies involving computational models.

Abstract

Despite differing from the human language processing mechanism in implementation and algorithms, current language models demonstrate remarkable human-like or surpassing language capabilities. Should computational language models be employed in studying the brain, and if so, when and how? To delve into this topic, this paper reviews efforts in using computational models for brain research, highlighting emerging trends. To ensure a fair comparison, the paper evaluates various computational models using consistent metrics on the same dataset. Our analysis reveals that no single model outperforms others on all datasets, underscoring the need for rich testing datasets and rigid experimental control to draw robust conclusions in studies involving computational models.
Paper Structure (13 sections, 4 equations, 3 figures, 1 table)

This paper contains 13 sections, 4 equations, 3 figures, 1 table.

Table of Contents

  1. Introduction
  2. Computational models and cognitive metrics
  3. Cognitive metrics
  4. Statistical language models
  5. Shallow Embedding Models
  6. Large Language Models
  7. Utilizing computational models in brain studies
  8. Quantifying cognitive load
  9. Formalizing syntactic processing
  10. Modelling linguistic Representations
  11. Verifying hypotheses for empirical validations
  12. Comparing models on a multimodal cognitive dataset
  13. Conclusion

Figures (3)

  • Figure 1: Statistical language models. (a) 3-gram language model, which estimates the word probability based on context. (b) structural language model, which incorporates syntax and parsing for understanding word interplay and sentence organization.
  • Figure 2: Shallow Embedding Models: (a) CBOW (Word2Vec) model that learn word embeddings by predicting target word based on context within a window, (b) RNN model that learning word embeddings by predicting next word based on all previous context.
  • Figure 3: Large Language Models: (a) BERT: Learns word embeddings through masked language modeling, predicting randomly masked words based on surrounding context. (b) GPT: Learns word embeddings via next word prediction, predicting the next word based on all preceding context.