Mathematical Model Suggests Seven Senses Optimal for Memory, Implications for AI and Robotics

Skoltech's Groundbreaking Memory Model

Scientists at Skoltech have developed a revolutionary mathematical model of memory that suggests seven senses might be optimal for maximizing memory capacity. This groundbreaking research, published in Scientific Reports, could have far-reaching implications for robot design, artificial intelligence, and our understanding of human memory 1

1

2

2

.

The Seven Senses Hypothesis

The study's co-author, Professor Nikolay Brilliantov of Skoltech AI, explains the core finding: "It appears that when each concept retained in memory is characterized in terms of seven features -- as opposed to, say, five or eight -- the number of distinct objects held in memory is maximized" 1

1

. This surprising conclusion suggests that humans, with our traditional five senses, might benefit from additional sensory inputs.

Memory Engrams and Conceptual Space

The research team modeled memory using "engrams," which are sparse ensembles of neurons that fire together across multiple brain regions. In this model, each engram represents an abstract object with multiple features corresponding to sensory inputs 1

1

2

2

. For example, the concept of a banana would be associated with its visual appearance, smell, taste, and other sensory characteristics.

Evolution of Memory and Learning

The model demonstrates how engrams evolve over time, becoming more focused or blurred depending on how often they are activated by external stimuli. This process simulates learning and forgetting through environmental interactions 1

1

. Professor Brilliantov notes, "We have mathematically demonstrated that the engrams in the conceptual space tend to evolve toward a steady state" 2

2

.

Optimal Dimensionality for Memory Capacity

The researchers discovered that a seven-dimensional conceptual space yields the highest number of distinct engrams stored in memory at steady state 1

1

2

2

. This finding led to the conclusion that seven senses might be optimal for maximizing memory capacity and, by extension, deepening overall understanding of the world.

Implications for AI and Robotics

While speculative in terms of human evolution, these findings could have significant practical applications in robotics and artificial intelligence 1

1

. The research suggests that designing AI systems with seven sensory inputs might optimize their memory capacity and information processing capabilities.

Robustness of the Seven-Sense Model

Importantly, the researchers found that the optimal number of seven dimensions persists regardless of the model's specific details, such as the properties of the conceptual space or the nature of sensory stimuli 1

1

2

2

. This robustness adds credibility to the findings and their potential applications.

Future Directions and Consciousness

The study of memory in living beings is closely tied to the enigmatic property of consciousness. As such, advancing theoretical models of memory, like the one developed by the Skoltech team, could provide valuable insights into the human mind and aid in the development of more sophisticated, human-like memory systems for AI agents 1

1

2

2

.

This research opens up exciting possibilities for future studies in neuroscience, cognitive science, and artificial intelligence, potentially reshaping our understanding of memory and sensory perception in both biological and artificial systems.