Biology subjects

Hassidim, A.

Publications and source records attributed to Hassidim, A..

5 recordsLinked to original sources

bioRxiv2024

Information-making processes in the speaker's brain drive human conversations forward

Natural human conversation is driven by the exchange of information-rich messages that surprise the listener and deviate from predictable context. While extensive research has characterized how the brain processes unexpected linguistic input during comprehension, the neural mechanisms underlying the generation of such information by the speaker remain poorly understood. In this study, we hypothesize that the speakers brain actively creates information through extra neural computation, transforming internal thoughts into novel, meaningful linguistic outputs. Utilizing a unique 24/7 dataset of continuous electrocorticography (ECoG) comprising approximately 100 hours of spontaneous natural conversations, we contrast the neural basis of speech production and comprehension within the same participants. Behaviorally, we find that speakers take longer pausing for an additional 100-150 milliseconds before producing information-rich, improbable words, even when controlling for word identity and frequency. Neurally, we provide converging evidence for a previously unreported process: generating information-rich words elicits significantly stronger neural activity (ERP) and enhanced neural encoding and decoding in language-related areas starting 100-500 ms before articulation. This pattern contrasts sharply with speech comprehension, where enhanced neural activity for predictable words occurs before onset, while responses to improbable words emerge only after onset as prediction errors. Furthermore, we demonstrate that large language models (LLMs) mirror this biological process, requiring deeper internal computation across layers to generate improbable versus probable words. Together, these results reveal that the speakers brain is not merely a transmission channel but an active generator of information, recruiting extra neural resources to create novel content that diverges from listener expectations.

neuroscience↗

bioRxiv2023

Deep speech-to-text models capture the neural basis of spontaneous speech in everyday conversations

Humans effortlessly use the continuous acoustics of speech to communicate rich linguistic meaning during everyday conversations. In this study, we leverage 100 hours (half a million words) of spontaneous open-ended conversations and concurrent high-quality neural activity recorded using electrocorticography (ECoG) to decipher the neural basis of real-world speech production and comprehension. Employing a deep multimodal speech-to-text model named Whisper, we develop encoding models capable of accurately predicting neural responses to both acoustic and semantic aspects of speech. Our encoding models achieved high accuracy in predicting neural responses in hundreds of thousands of words across many hours of left-out recordings. We uncover a distributed cortical hierarchy for speech and language processing, with sensory and motor regions encoding acoustic features of speech and higher-level language areas encoding syntactic and semantic information. Many electrodes--including those in both perceptual and motor areas--display mixed selectivity for both speech and linguistic features. Notably, our encoding model reveals a temporal progression from language-to-speech encoding before word onset during speech production and from speech-to-language encoding following word articulation during speech comprehension. This study offers a comprehensive account of the unfolding neural responses during fully natural, unbounded daily conversations. By leveraging a multimodal deep speech recognition model, we highlight the power of deep learning for unraveling the neural mechanisms of language processing in real-world contexts.

neuroscience↗

bioRxiv2022

Correspondence between the layered structure of deep language models and temporal structure of natural language processing in the human brain

Deep language models (DLMs) provide a novel computational paradigm for how the brain processes natural language. Unlike symbolic, rule-based models described in psycholinguistics, DLMs encode words and their context as continuous numerical vectors. These "embeddings" are constructed by a sequence of computations organized in "layers" to ultimately capture surprisingly sophisticated representations of linguistic structures. How does this layered hierarchy map onto the human brain during natural language comprehension? In this study, we used electrocorticography (ECoG) to record neural activity in language areas along the superior temporal gyrus and inferior frontal gyrus while human participants listened to a 30-minute spoken narrative. We supplied this same narrative to a high-performing DLM (GPT2-XL) and extracted the contextual embeddings for each word in the story across all 48 layers of the model. We next trained a set of linear encoding models to predict the temporally-evolving neural activity from the embeddings at each layer. We found a striking correspondence between the layer-by-layer sequence of embeddings from GPT2-XL and the temporal sequence of neural activity in language areas. In addition, we found evidence for the gradual accumulation of recurrent information along the linguistic processing hierarchy. However, we also noticed additional neural processes in the brain, but not in DLMs, during the processing of surprising (unpredictable) words. These findings point to a connection between human language processing and DLMs where the layer-by-layer accumulation of contextual information in DLM embeddings matches the temporal dynamics of neural activity in high-order language areas.

neuroscience↗

bioRxiv2022

Brain embeddings with shared geometry to artificial contextual embeddings, as a code for representing language in the human brain

Contextual embeddings, derived from deep language models (DLMs), provide a continuous vectorial representation of language. This embedding space differs fundamentally from the symbolic representations posited by traditional psycholinguistics. Do language areas in the human brain, similar to DLMs, rely on a continuous embedding space to represent language? To test this hypothesis, we densely recorded the neural activity in the Inferior Frontal Gyrus (IFG, also known as Brocas area) of three participants using dense intracranial arrays while they listened to a 30-minute podcast. From these fine-grained spatiotemporal neural recordings, we derived for each patient a continuous vectorial representation for each word (i.e., a brain embedding). Using stringent, zero-shot mapping, we demonstrated that brain embeddings in the IFG and the DLM contextual embedding space have strikingly similar geometry. This shared geometry allows us to precisely triangulate the position of unseen words in both the brain embedding space (zero-shot encoding) and the DLM contextual embedding space (zero-shot decoding). The continuous brain embedding space provides an alternative computational framework for how natural language is represented in cortical language areas.

neuroscience↗

bioRxiv2020

Thinking ahead: prediction in context as a keystone of language in humans and machines

Departing from traditional linguistic models, advances in deep learning have resulted in a new type of predictive (autoregressive) deep language models (DLMs). Using a self-supervised next-word prediction task, these models are trained to generate appropriate linguistic responses in a given context. We provide empirical evidence that the human brain and autoregressive DLMs share three fundamental computational principles as they process natural language: 1) both are engaged in continuous next-word prediction before word-onset; 2) both match their pre-onset predictions to the incoming word to calculate post-onset surprise (i.e., prediction error signals); 3) both represent words as a function of the previous context. In support of these three principles, our findings indicate that: a) the neural activity before word-onset contains context-dependent predictive information about forthcoming words, even hundreds of milliseconds before the words are perceived; b) the neural activity after word-onset reflects the surprise level and prediction error; and c) autoregressive DLM contextual embeddings capture the neural representation of context-specific word meaning better than arbitrary or static semantic embeddings. Together, our findings suggest that autoregressive DLMs provide a novel and biologically feasible computational framework for studying the neural basis of language.

neuroscience↗