Probabilistic numerical discrimination in mice.

Area of Science:

• Cognitive neuroscience
• Animal behavior
• Decision-making

Background:

• Animals, including humans, can distinguish between quantities like time and number, but with inherent precision limits due to internal uncertainty.
• Previous research indicates that temporal categorization can be adaptively modulated by incorporating probabilistic information into time-based decisions.

Purpose of the Study:

• To investigate whether mice can integrate external probabilistic information into their numerosity-based judgments.
• To explore the parallels between temporal and numerical decision-making under probabilistic influences in mice.

Main Methods:

• Mice were trained on a task where rewards were contingent on a specific number of lever presses (e.g., 10 or 20).
• The relative frequencies of different trial types (numerosity conditions) were manipulated across experimental sessions to introduce probabilistic information.
• Behavioral performance was analyzed using models that considered response costs to evaluate adaptive decision-making.

Main Results:

• Mice demonstrated the ability to adaptively modulate their count-based decisions in response to experienced probabilistic contingencies.
• The observed adjustments in decision-making aligned with predictions derived from optimality principles.
• This study provides the first evidence of mice integrating exogenous probabilistic information into numerosity judgments.

Conclusions:

• Mice can dynamically adjust their numerical judgments based on learned probabilistic information from their environment.
• The findings suggest a shared mechanism for integrating probabilistic cues across different sensory domains (time and number) in mice.
• This research advances our understanding of the flexibility and adaptability of numerical cognition in non-human animals.