This study investigates the neural representation of semantic information during language comprehension at the single-cell level. By recording from single neurons in the left language-dominant prefrontal cortex while participants listened to semantically diverse sentences and naturalistic stories, the researchers discovered that individual neurons selectively respond to specific word meanings and can distinguish words from nonwords. These neurons' activities are highly dynamic, reflecting the meanings of words based on their specific sentence contexts and independent of their phonetic form. The combined activity patterns of these neurons can accurately predict the semantic domains of words, even when tested with different linguistic materials. The findings suggest that these neurons robustly represent semantic information and can rapidly map it onto the population's response patterns during speech. Additionally, the neurons' responses are context-dependent, adapting to the specific sentences in which words are heard, even when the words are phonetically identical. This study provides insights into how individual neurons represent and process semantic information during language comprehension, highlighting the potential for modality-independent semantic representations and the possibility of similar processes in other parts of the brain.This study investigates the neural representation of semantic information during language comprehension at the single-cell level. By recording from single neurons in the left language-dominant prefrontal cortex while participants listened to semantically diverse sentences and naturalistic stories, the researchers discovered that individual neurons selectively respond to specific word meanings and can distinguish words from nonwords. These neurons' activities are highly dynamic, reflecting the meanings of words based on their specific sentence contexts and independent of their phonetic form. The combined activity patterns of these neurons can accurately predict the semantic domains of words, even when tested with different linguistic materials. The findings suggest that these neurons robustly represent semantic information and can rapidly map it onto the population's response patterns during speech. Additionally, the neurons' responses are context-dependent, adapting to the specific sentences in which words are heard, even when the words are phonetically identical. This study provides insights into how individual neurons represent and process semantic information during language comprehension, highlighting the potential for modality-independent semantic representations and the possibility of similar processes in other parts of the brain.