Task interference as a neuronal basis for the cost of cognitive flexibility

Cheng Xue,Sol K. Markman,Ruoyi Chen,Lily E. Kramer,Marlene R. Cohen

DOI: https://doi.org/10.1101/2024.03.04.583375

2024-03-06

Abstract:Humans and animals have an impressive ability to juggle multiple tasks in a constantly changing environment. This flexibility, however, leads to decreased performance under uncertain task conditions. Here, we combined monkey electrophysiology, human psychophysics, and artificial neural network modeling to investigate the neuronal mechanisms of this performance cost. We developed a behavioural paradigm to measure and influence participants’ decision-making and perception in two distinct perceptual tasks. Our data revealed that both humans and monkeys, unlike an artificial neural network trained for the same tasks, make less accurate perceptual decisions when the task is uncertain. We generated a mechanistic hypothesis by comparing this neural network trained to produce correct choices with another network trained to replicate the participants’ choices. We hypothesized, and confirmed with further behavioural, physiological, and causal experiments, that the cost of task flexibility comes from what we term task interference. Under uncertain conditions, interference between different tasks causes errors because it results in a stronger representation of irrelevant task features and entangled neuronal representations of different features. Our results suggest a tantalizing, general hypothesis: that cognitive capacity limitations, both in health and disease, stem from interference between neural representations of different stimuli, tasks, or memories.

Neuroscience

What problem does this paper attempt to address?

### The problems the paper attempts to solve

The paper aims to explore the cost of cognitive flexibility and its neural mechanisms. Specifically, the authors studied why the cognitive performance of humans and animals declines under uncertain task conditions. They combined electrophysiological experiments on monkeys, psychophysical experiments on humans, and artificial neural network modeling to investigate the neural basis of this performance decline.

### Main problems

1. **Cost of cognitive flexibility**:
- Why do the perceptual decisions of humans and animals become less accurate under uncertain task conditions?
- What are the neural mechanisms of this performance decline?

2. **Task Interference**:
- How does task interference under uncertain task conditions lead to errors?
- Does task interference lead to stronger representations of irrelevant task features and entanglement of neural representations of different features?

### Research methods

- **Behavioral experiments**:
- Humans and monkeys participated in a two - feature visual discrimination task, which required distinguishing between two visual features (such as spatial frequency and position) and ignoring another feature.
- By changing the certainty of the task (high or low), the behavioral performance of the participants was observed.

- **Electrophysiological experiments**:
- The neuronal activities in the primary visual cortex (V1) of monkeys were recorded, and the neural representations under different task certainty conditions were analyzed.

- **Artificial neural network modeling**:
- Two different recurrent neural networks (RNNs) were trained: one was the "Correct Choice Network", and the other was the "Monkey Choice Network".
- The performance of these two networks under different task certainty conditions was compared, and a hypothesis about the mechanism of task interference was generated.

### Conclusions

The authors found that the cost of cognitive flexibility mainly comes from task interference. Under uncertain task conditions, task interference will lead to stronger representations of irrelevant task features and entanglement of neural representations of different features, thus affecting the accuracy of perceptual decisions. This finding provides a new neural mechanism explanation for understanding the cost of cognitive flexibility and a theoretical basis for further research.