analiza aiqbench
This research introduces generative causal testing (GCT), a framework that uses large language models to generate testable hypotheses about brain selectivity and validates them against fMRI data. By bridging the gap between opaque predictive models and formal scientific theories, GCT successfully explains neural responses in specific cortical regions using LLM-generated stimuli. The study demonstrates how AI can move beyond correlation to provide causal explanations for complex cognitive processes like language processing.
Generative causal testing to bridge data-driven models and scientific theories in language neuroscience
arXiv:2410.00812v3 Announce Type: replace Abstract: Representations from large language models are highly effective at predicting BOLD fMRI responses to language stimuli. However, these representations are largely opaque: it is unclear what features of the language stimulus drive the response in each brain area. We present generative causal testing (GCT), a framework for generating concise explanations of language selectivity in the brain from predictive models and then testing those explanations in follow-up experiments using LLM-generated stimuli.This approach is successful at explaining selectivity both in individual voxels and cortical regions of interest (ROIs), including newly identified microROIs in prefrontal cortex. We show that explanatory accuracy is closely related to the predictive power and stability of the underlying predictive models. Finally, we show that GCT can dissect fine-grained differences between brain areas with similar functional selectivity. These results demonstrate that LLMs can be used to bridge the widening gap between data-driven models and formal scientific theories.
Abstract:Representations from large language models are highly effective at predicting BOLD fMRI responses to language stimuli. However, these representations are largely opaque: it is unclear what features of the language stimulus drive the response in each brain area. We present generative causal testing (GCT), a framework for generating concise explanations of language selectivity in the brain from predictive models and then testing those explanations in follow-up experiments using LLM-generated this http URL approach is successful at explaining selectivity both in individual voxels and cortical regions of interest (ROIs), including newly identified microROIs in prefrontal cortex. We show that explanatory accuracy is closely related to the predictive power and stability of the underlying predictive models. Finally, we show that GCT can dissect fine-grained differences between brain areas with similar functional selectivity. These results demonstrate that LLMs can be used to bridge the widening gap between data-driven models and formal scientific theories.
Computation and Language (cs.CL); Neurons and Cognition (q-bio.NC)
arXiv:2410.00812 [cs.CL]
(or arXiv:2410.00812v3 [cs.CL] for this version)
Text integral sub licența deschisă a publicației · 166 cuvinte
