Evidence of a predictive coding hierarchy in the human brain listening to speech
Bibliographic Data
| ID | 4681977 |
|---|---|
| Authors | Charlotte Caucheteux (0000-0002-3965-3143, École Polytechnique, corresponding author), Alexandre Gramfort (0000-0001-9791-4404, École Polytechnique), Jean-Remi King (0000-0002-2121-170X, Centre National de la Recherche Scientifique, corresponding author) |
| Year | 2023 |
| Volume | 7 |
| Issue | 3 |
| Pages | 430-441 |
| Publication date | 2023-03-02 |
| Peer Reviewed | Yes |
| Open Access | Yes |
| Type | ARTICLE |
| Venue | Nature Human Behaviour (JOURNAL) |
| Journal identifiers | ISSN: 2397-3374 • E-ISSN: 2397-3374 |
| Publisher | Springer Science and Business Media LLC (PUBLISHER) |
| DOI | 10.1038/s41562-022-01516-2 |
| PMID | 36864133 |
| OpenAlex | W4322766928 |
| Language | EN |
| Citations received | 14 |
| References cited | 69 |
Considerable progress has recently been made in natural language processing: deep learning algorithms are increasingly able to generate, summarize, translate and classify texts. Yet, these language models still fail to match the language abilities of humans. Predictive coding theory offers a tentative explanation to this discrepancy: while language models are optimized to predict nearby words, the human brain would continuously predict a hierarchy of representations that spans multiple timescales. To test this hypothesis, we analysed the functional magnetic resonance imaging brain signals of 304 participants listening to short stories. First, we confirmed that the activations of modern language models linearly map onto the brain responses to speech. Second, we showed that enhancing these algorithms with predictions that span multiple timescales improves this brain mapping. Finally, we showed that these predictions are organized hierarchically: frontoparietal cortices predict higher-level, longer-range and more contextual representations than temporal cortices. Overall, these results strengthen the role of hierarchical predictive coding in language processing and illustrate how the synergy between neuroscience and artificial intelligence can unravel the computational bases of human cognition
Active listening · Cognition · Cognitive psychology · Cognitive science · Computational model · Functional magnetic resonance imaging · Hierarchy · Human brain · Language model · Natural language processing · Predictive coding · Action Observation and Synchronization · Communication · Computer Science · Neurobiology of Language and Bilingualism · Neuroscience · Psychology · Topic Modeling · Artificial Intelligence
Large-scale benchmark yields no evidence that language model surprisal explains syntactic disambiguation difficulty
Trauma and surprise in free energy models of the mind
The Simulative Role of Neural Language Models in Brain Language Processing
From Voice to Self
Incremental alternative sampling as a lens into the temporal and representational resolution of linguistic prediction
Decoding and controlling emotion in LLMs through human-aligned representational geometry with enhanced interpretability
Temporal Dynamics of Intraparietal Sulcus Activation in Response to Tonal Uncertainty
Guessing at Ghosts in the Machine
Dynamical systems, literary theory, and the computational modelling of narrative
Conversational content is organized across multiple timescales in the brain
A unified acoustic-to-speech-to-language embedding space captures the neural basis of natural language processing in everyday conversations
Driving and suppressing the human language network using large language models
Two-dimensional neural geometry underpins hierarchical organization of sequence in human working memory
Neural populations in the language network differ in the size of their temporal receptive windows
SciPy 1.0
Transformers
An interactive activation model of context effects in letter perception
Predictive coding in the visual cortex
The cortical organization of speech processing
Cortical tracking of hierarchical linguistic structures in connected speech
Bart
Distributed Hierarchical Processing in the Primate Cerebral Cortex
Predictive coding under the free-energy principle
Natural speech reveals the semantic maps that tile human cerebral cortex
Neurocomputational Models of Language Processing
| Unique citing works | 14 |
|---|---|
| Citations per year | 7 |
| Citation span | 2024 - 2026 (3) |
| Citation velocity | current |
| Highly cited | No |
| Citation types | Neutral: 13 |