Human brain

Understanding

• Key-value memory in the brain
• The cost of brain state transitions

Brain mapping

• 2024-05: A petavoxel fragment of human cerebral cortex reconstructed at nanoscale resolution (media summary)
• 2024-10: Neuronal wiring diagram of an adult brain (media summary); 140,000 neurons in fruit fly brain
• 2024-12: A roadmap to scale connectomics to entire mammalian brains
• 2025-04: Functional connectomics reveals general wiring rule in mouse visual cortex (media summary)
• 2025-08: Light-microscopy-based connectomic reconstruction of mammalian brain tissue (blog)

Brain signal decoding

• 2022-11: High-resolution image reconstruction with latent diffusion models from human brain activity
• 2023-08: Music can be reconstructed from human auditory cortex activity using nonlinear decoding models (intracranial EEG)
• 2023-09: DeWave: Discrete EEG Waves Encoding for Brain Dynamics to Text Translation (external EEG)
• 2023-09: BrainLM: A foundation model for brain activity recordings
• 2023-10: Toward a real-time decoding of images from brain activity (MEG)
• 2024-06: PAM: Predictive Attention Mechanism for Neural Decoding of Visual Perception
• 2024-07: Scaling Law in Neural Data: Non-Invasive Speech Decoding with 175 Hours of EEG Data
• 2024-12: Predicting Human Brain States with Transformer
• 2025-02: Meta: Brain-to-Text Decoding: A Non-invasive Approach via Typing
• 2025-02: Meta: From Thought to Action: How a Hierarchy of Neural Dynamics Supports Language Production
• 2025-03: Google: Deciphering language processing in the human brain through LLM representations
• 2025-03: A streaming brain-to-voice neuroprosthesis to restore naturalistic communication

Computational Analysis

Comparison to computer

• How (and Why) to Think that the Brain is Literally a Computer
• Contextual feature extraction hierarchies converge in large language models and the brain (LLMs are becoming more brain-like as they advance)

Biological vs. artificial neuron

• Single cortical neurons as deep artificial neural networks: Each biological neuron can be simulated using DNN of 5-8 layers
• Mapping Biological Neuron Dynamics into an Interpretable Two-layer Artificial Neural Network

Data processing

• How Much the Eye Tells the Brain
• Representational geometry: integrating cognition, computation, and the brain
• Language is primarily a tool for communication rather than thought
• The Unbearable Slowness of Being: Why do we live at 10 bits/s? (preprint)

Extract manifold/geometry

• Selection of experience for memory by hippocampal sharp wave ripples

Comparisons

• 2024-05: Are Biological Systems More Intelligent Than Artificial Intelligence?
• 2025-03: Google: Deciphering language processing in the human brain through LLM representations
  • 2022-03: Shared computational principles for language processing in humans and deep language models
  • 2024-03: Alignment of brain embeddings and artificial contextual embeddings in natural language points to common geometric patterns
  • 2025-03: A unified acoustic-to-speech-to-language embedding space captures the neural basis of natural language processing in everyday conversations

Simulate Brain

• 2023-09: The Digital Twin Brain: A Bridge between Biological and Artificial Intelligence
• 2024-12: Simulation and assimilation of the digital human brain (preprint, code)
• 2024-12: Predicting Human Brain States with Transformer

Bio-brain Inspirations for AI

• 2025-01: TopoNets: High Performing Vision and Language Models with Brain-Like Topography

See Also

• LLM Simulate Humans