The brain-AI convergence: Predictive and generative world models for general-purpose computation

2025/12/02 by Shogo Ohmae, Ohmae, Shogo, Keiko Ohmae +1
Engineering · Neuroscience · #Advanced Memory and Neural Computing #Ferroelectric and Negative Capacitance Devices #EEG and Brain-Computer Interfaces

paper · pdf · doi:10.48550/arxiv.2512.02419

Abstract

Recent advances in general-purpose AI systems with attention-based transformers offer a potential window into how the neocortex and cerebellum, despite their relatively uniform circuit architectures, give rise to diverse functions and, ultimately, to human intelligence. This Perspective provides a cross-domain comparison between the brain and AI that goes beyond the traditional focus on visual processing, adopting the emerging perspecive of world-model-based computation. Here, we identify shared computational mechanisms in the attention-based neocortex and the non-attentional cerebellum: both predict future world events from past inputs and construct internal world models through prediction-error learning. These predictive world models are repurposed for seemingly distinct functions -- understanding in sensory processing and generation in motor processing -- enabling the brain to achieve multi-domain capabilities and human-like adaptive intelligence. Notably, attention-based AI has independently converged on a similar learning paradigm and world-model-based computation. We conclude that these shared mechanisms in both biological and artificial systems constitute a core computational foundation for realizing diverse functions including high-level intelligence, despite their relatively uniform circuit structures. Our theoretical insights bridge neuroscience and AI, advancing our understanding of the computational essence of intelligence.

Citations

• Beyond Induction Heads: In-Context Meta Learning Induces Multi-Phase Circuit Emergence
• Which Attention Heads Matter for In-Context Learning?
• The Topos of Transformer Networks
• PIM-GPT: A Hybrid Process-in-Memory Accelerator for Autoregressive Transformers
• One Step of Gradient Descent is Provably the Optimal In-Context Learner with One Layer of Linear Self-Attention
• Trained Transformers Learn Linear Models In-Context
• A Survey of Large Language Models
• World Models and Predictive Coding for Cognitive and Developmental Robotics: Frontiers and Challenges
• Transformers learn in-context by gradient descent
• What learning algorithm is in-context learning? Investigations with linear models
• What Can Transformers Learn In-Context? A Case Study of Simple Function Classes
• A Generalist Agent
• Do As I Can, Not As I Say: Grounding Language in Robotic Affordances
• BNAI, NO-TOKEN, and MIND-UNITY: Pillars of a Systemic Revolution in Artificial Intelligence
• Masked Autoencoders Are Scalable Vision Learners
• From internal models toward metacognitive AI
• Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity
• Language Models are Few-Shot Learners
• 50 years since the Marr, Ito, and Albus models of the cerebellum
• Scaling Laws for Neural Language Models
• Linear Mode Connectivity and the Lottery Ticket Hypothesis
• Cortical microcircuits as gated-recurrent neural networks
• Attention Is All You Need
• Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation
• Deep Predictive Coding Networks for Video Prediction and Unsupervised Learning
• Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning
• Understanding Neural Networks Through Deep Visualization
• Learning both Weights and Connections for Efficient Neural Networks
• Sequence to Sequence Learning with Neural Networks
• Visualizing and Understanding Convolutional Networks
• ImageNet classification with deep convolutional neural networks
• The Attention System of the Human Brain: 20 Years After
• Neural Basis of Reinforcement Learning and Decision Making
• Building high-level features using large scale unsupervised learning
• Finding Structure in Time
• Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position
• A hierarchical neural-network model for control and learning of voluntary movement
• An Integrative Theory of Prefrontal Cortex Function
• Neural Mechanisms of Selective Visual Attention
• The Attention System of the Human Brain

Related

• The brain versus AI: World-model-based versatile circuit computation underlying diverse functions in the neocortex and cerebellum
• From generative AI to the brain: five takeaways
• A brain basis of dynamical intelligence for AI and computational neuroscience
• Cognitive computational neuroscience
• Alignment between Brains and AI: Evidence for Convergent Evolution across Modalities, Scales and Training Trajectories