Neural Network Can Weigh Options Like a Human – New Paper

Bill Loguidice — Tue, 16 Jul 2024 15:28:38 +0000

In a paper in the journal Nature Human Behaviour, “The Neural Network RTNet Exhibits the Signatures of Human Perceptual Decision-Making,” a team from the Georgia Tech School of Psychology shows that while convolutional neural networks show promise as models of biological vision, they have limited applicability to human decision-making. Instead, the team has developed a new neural network, RTNet, that generates stochastic decisions and human-like response time (RT) distributions. Comprehensive tests showed RTNet reproduces all foundational features of human accuracy, RT, and confidence, and does so better than all current alternatives. The abstract follows.

Neural networks can help computers make better decisions independently.

Abstract

Convolutional neural networks show promise as models of biological vision. However, their decision behaviour, including the facts that they are deterministic and use equal numbers of computations for easy and difficult stimuli, differs markedly from human decision-making, thus limiting their applicability as models of human perceptual behaviour. Here we develop a new neural network, RTNet, that generates stochastic decisions and human-like response time (RT) distributions. We further performed comprehensive tests that showed RTNet reproduces all foundational features of human accuracy, RT and confidence and does so better than all current alternatives. To test RTNet’s ability to predict human behaviour on novel images, we collected accuracy, RT and confidence data from 60 human participants performing a digit discrimination task. We found that the accuracy, RT and confidence produced by RTNet for individual novel images correlated with the same quantities produced by human participants. Critically, human participants who were more similar to the average human performance were also found to be closer to RTNet’s predictions, suggesting that RTNet successfully captured average human behaviour. Overall, RTNet is a promising model of human RTs that exhibits the critical signatures of perceptual decision-making.

The post Neural Network Can Weigh Options Like a Human – New Paper appeared first on fullSTEAMahead365.

Boredom can lead to negative behavior in new study

Bill Loguidice — Thu, 23 May 2024 14:55:08 +0000

In a new paper from Motivation and Emotion via Springer Nature’s Springer Link, boredom can lead to non-suicidal self-injury behaviors, even when more positive alternatives are available. In the study, boredom increased the selection of negative stimuli and negative sounds versus situations that induced anger or even were more neutral. You can read more about the causal effect of boredom and the self-administration of aversive stimuli in the presence of a positive alternative at the link. The abstract follows.

Even with better alternatives readily available, being bored can lead to poorer choices.

Abstract

Previous studies demonstrated that boredom leads to increased self-administration of painful electric stimulation, a proxy for non-suicidal self-injury (NSSI). However, in these experimental studies, participants had no behavioral alternatives besides electric stimulation to break the monotony. A first aim in the current experimental study was to examine whether boredom still leads to self-administering aversive stimuli when positive behavioral alternatives are present. This boredom effect was furthermore compared to an anger induction. The second aim was to examine whether history of NSSI and negative urgency (i.e., the tendency to engage in impulsive action in response to negative emotions) positively moderate the link between boredom and self-administered aversive stimuli. In a between-subjects design using college students (N = 129), participants were randomly assigned to one of three emotion induction writing tasks (i.e., boredom, anger, neutral), during which frequency of self-administered positive (chirping birds) and aversive (screaming pig) sounds was measured. The latter was used as a proxy for NSSI behavior. Results showed that boredom led to increased selection of aversive sounds compared to the neutral and anger conditions, despite the presence of a positive alternative (i.e., positive sounds). No difference in frequency of selecting the aversive sounds was observed between the anger and neutral condition. Neither history of NSSI nor negative urgency moderated the effect of condition on self-administered aversive stimuli. The current results tentatively support a causal and specific link between boredom and NSSI, and warrant further examination of the role of boredom in maladaptive behaviors such as NSSI.

The post Boredom can lead to negative behavior in new study appeared first on fullSTEAMahead365.

paper Archives - fullSTEAMahead365

Neural Network Can Weigh Options Like a Human – New Paper

Abstract

Boredom can lead to negative behavior in new study

Abstract