Tracking Emotions: Intrinsic Motivation Grounded on Multi - Level Prediction Error Dynamics

Schillaci, Guido; Ciria, Alejandra; Lara, Bruno

doi:10.1109/icdl-epirob48136.2020.9278106

Cited by 12 publications

(20 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This scheme supports development of hierarchical information processing by adequately setting the timescale of each layer [14], [21]. This approach is considered as analogous to [24], [25].…”

Section: B Overview Of Pv-rnnmentioning

confidence: 99%

Leading or Following? Dyadic Robot Imitative Interaction Using the Active Inference Framework

Wirkuttis

Tani

2021

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

Section: B Overview Of Pv-rnnmentioning

confidence: 99%

Leading or Following? Dyadic Robot Imitative Interaction Using the Active Inference Framework

Wirkuttis

Tani

2021

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

“…Different prominent scholars describe emotions as a form of nonconceptual monitoring of coping performance in our interaction with the world (Frijda, 2006;Reisenzein, 2009). We just propose that prediction error dynamics are the form of feedback on the system's own functioning required to spell out those accounts in computational specificity (Hesp et al, 2021;Joffily & Coricelli, 2013;Schillaci et al, 2020).…”

Section: Learning Dynamics and Valuationmentioning

confidence: 97%

Preferences need inferences: Learning, valuation, and curiosity in aesthetic experience

Cruys¹,

Bervoets²,

Moors³

2021

Preprint

View full text Add to dashboard Cite

More than 40 years ago, pioneering social psychologist Robert Zajonc (1980) published his seminal work titled “Preferences need no inferences” in which he argued for the primacy of affect over cognition. Affective evaluation (the preference) comes first, he claimed, and only then do cognitive processes (the inferences) kick in. The view is untenable in light of recent predictive processing accounts of the mind, which hold that all mental functioning is built from (approximate) Bayesian inference. It casts perception, action, and learning as inference but, perhaps counterintuitively, valuation too. We discuss how valuation —understood as the process of how we come to value, prefer or like things— emerges as a function of learning and inference, and how this conception may help us resolve traditional conundrums in the science of aesthetic experience, such as the nature of the "beholder's share", the link between curiosity and appreciation, Keats' "negative capability" and the tension between the mere exposure principle and the goldilocks (optimal level) principle.

show abstract

“…Here, selecting the best task among solicitations is based on their associated expected error reduction rate. This rate is learned and constantly updated during situated action cycles, being directly linked to the current competence of the agent to achieve the desired outcome (for an implementation see Schillaci et al, 2020a ).…”

Section: Interactionist Model Of Contextsmentioning

confidence: 99%

“…Different time windows of prediction error monitoring, starting from being very brief to relatively long, produce different patterns of emotional experience, as well as a different sensitivity to meaningful changes in the error reduction rate (Carver and Scheier, 1990 ). Recently, it has been suggested that the size of this time window should change dynamically according to ‘how well or bad things are going' with respect to the expected progress (Schillaci et al, 2020a , b ). Thus, when the error rate constantly decreases, meaning the agent is doing well on the task execution, the need for error monitoring diminishes.…”

Section: Interactionist Model Of Contextsmentioning

confidence: 99%

Grounding Context in Embodied Cognitive Robotics

Valenzo

Ciria

Schillaci³

et al. 2022

Front. Neurorobot.

Self Cite

View full text Add to dashboard Cite

Biological agents are context-dependent systems that exhibit behavioral flexibility. The internal and external information agents process, their actions, and emotions are all grounded in the context within which they are situated. However, in the field of cognitive robotics, the concept of context is far from being clear with most studies making little to no reference to it. The aim of this paper is to provide an interpretation of the notion of context and its core elements based on different studies in natural agents, and how these core contextual elements have been modeled in cognitive robotics, to introduce a new hypothesis about the interactions between these contextual elements. Here, global context is categorized as agent-related, environmental, and task-related context. The interaction of their core elements, allows agents to first select self-relevant tasks depending on their current needs, or for learning and mastering their environment through exploration. Second, to perform a task and continuously monitor its performance. Third, to abandon a task in case its execution is not going as expected. Here, the monitoring of prediction error, the difference between sensorimotor predictions and incoming sensory information, is at the core of behavioral flexibility during situated action cycles. Additionally, monitoring prediction error dynamics and its comparison with the expected reduction rate should indicate the agent its overall performance on executing the task. Sensitivity to performance evokes emotions that function as the driving element for autonomous behavior which, at the same time, depends on the processing of the interacting core elements. Taking all these into account, an interactionist model of contexts and their core elements is proposed. The model is embodied, affective, and situated, by means of the processing of the agent-related and environmental core contextual elements. Additionally, it is grounded in the processing of the task-related context and the associated situated action cycles during task execution. Finally, the model proposed here aims to guide how artificial agents should process the core contextual elements of the agent-related and environmental context to give rise to the task-related context, allowing agents to autonomously select a task, its planning, execution, and monitoring for behavioral flexibility.

show abstract

Tracking Emotions: Intrinsic Motivation Grounded on Multi - Level Prediction Error Dynamics

Cited by 12 publications

References 20 publications

Leading or Following? Dyadic Robot Imitative Interaction Using the Active Inference Framework

Leading or Following? Dyadic Robot Imitative Interaction Using the Active Inference Framework

Preferences need inferences: Learning, valuation, and curiosity in aesthetic experience

Grounding Context in Embodied Cognitive Robotics

Contact Info

Product

Resources

About