The field of cognitive neuroscience and language processing is moving towards a deeper understanding of the relationship between brain activity and language comprehension. Recent studies have focused on improving the alignment between language models and human brain activity, with a particular emphasis on the integration of associative memory and the development of more sophisticated language models. The use of transformer-based models has shown promise in explaining predictability effects in eye movement behaviors, while the integration of associative memory has improved the alignment between language models and brain activity in regions related to associative memory processing. Additionally, there has been significant progress in the development of models that can decode EEG signals into text, with the introduction of new models such as the R1 Translator, which has achieved state-of-the-art performance in EEG-to-text translation tasks. Furthermore, studies have demonstrated the effectiveness of using large-scale language and speech models to reconstruct neural activity recordings captured during speech production, and have highlighted the importance of examining overt speech production paradigms in the development of brain-computer interfaces. Noteworthy papers include: The study on Modeling cognitive processes of natural reading with transformer-based Language Models, which evaluated the performance of transformer-based models in explaining predictability effects in eye movement behaviors. The paper on Improve Language Model and Brain Alignment via Associative Memory, which proposed a new approach to improving the alignment between language models and human brain activity by integrating associative memory. The study on EEG-to-Text Translation: A Model for Deciphering Human Brain Activity, which introduced a new model, the R1 Translator, for EEG-to-text translation tasks. The paper on Recreating Neural Activity During Speech Production with Language and Speech Model Embeddings, which demonstrated the effectiveness of using large-scale language and speech models to reconstruct neural activity recordings captured during speech production.