Toggle light / dark theme

Abstract: We explore the emergence of intelligent behavior in artificial systems by investigating how the complexity of rule-based systems influences the capabilities of models trained to predict these rules. Our study focuses on elementary cellular automata (ECA), simple yet powerful one-dimensional systems that generate behaviors ranging from trivial to highly complex. By training distinct Large Language Models (LLMs) on different ECAs, we evaluated the relationship between the complexity of the rules’ behavior and the intelligence exhibited by the LLMs, as reflected in their performance on downstream tasks. Our findings reveal that rules with higher complexity lead to models exhibiting greater intelligence, as demonstrated by their performance on reasoning and chess move prediction tasks. Both uniform and periodic systems, and often also highly chaotic systems, resulted in poorer downstream performance, highlighting a sweet spot of complexity conducive to intelligence. We conjecture that intelligence arises from the ability to predict complexity and that creating intelligence may require only exposure to complexity.

From: Shiyang Zhang [view email].

Summary: Researchers have identified a protein complex, TrkC-PTPσ, that plays a key role in the structural organization of synapses in the brain, impacting cognitive behaviors. By studying this complex, scientists uncovered how it regulates synaptic protein phosphorylation, essential for healthy brain function. Disruptions in this protein complex led to anxiety-like behaviors in mice, providing insights into mental health conditions like anxiety and autism.

The study sheds light on synaptic mechanisms that could help develop new therapeutic strategies. These findings advance our understanding of synapse function and its role in cognitive disorders, bringing hope for targeted treatment options in the future.