The probabilistic and dynamic nature of perception in human generalization behavior

Rafal Tekreeti
Jun 13
2 min read

By: Kenny Yu, Wolf Vanpaemel, Francis Tuerlinckx, Jonas Zaman

This research introduces a computational model that integrates perceptual and generalization theories to explore how humans transfer learned knowledge to new situations.

Analyzing data from Pavlovian fear conditioning experiments, the study reveals significant individual differences in perceptual processes, which directly impact how fear experiences generalize.

Figure 1: Conceptualizing the Mental Representation of Stimuli. This figure illustrates different ways the human mind might represent physical stimuli and how these representations change or remain stable over time. Panel (a) "Static mental mapping" depicts a fixed relationship between physical size and perceived size that doesn't change over time. Panel (b) "Dynamic mapping" shows that this relationship can evolve. Panel (c) "Probabilistic mapping," which aligns with our research, suggests that mental representations are not fixed points but are probabilistic and dynamic, meaning our perception of the same stimulus can vary and change with experience. The perceived distance between a learned context and a novel one is thus also dynamic and probabilistic. — **Figure 1:** Conceptualizing the Mental Representation of Stimuli. This figure illustrates different ways the human mind might represent physical stimuli and how these representations change or remain stable over time. Panel (a) "Static mental mapping" depicts a fixed relationship between physical size and perceived size that doesn't change over time. Panel (b) "Dynamic mapping" shows that this relationship can evolve. Panel (c) "Probabilistic mapping," which aligns with our research, suggests that mental representations are not fixed points but are probabilistic and dynamic, meaning our perception of the same stimulus can vary and change with experience. The perceived distance between a learned context and a novel one is thus also dynamic and probabilistic.

The findings highlight that perception is not static but varies between individuals and within individuals over time. By modeling perception as dynamic probability distributions, the research uncovers the combined role of perceptual and generalization mechanisms in producing behavior.

This work emphasizes the probabilistic perceptual foundations underlying individual differences in generalization, moving beyond traditional theories that often overlook the dynamic and uncertain nature of mental representations.

Figure 2: Computational Model Framework Overview. This diagram outlines the architecture of our computational model. It shows how perceptual data is processed by a state-space model (using a Bayesian-like process) to inform a "Model-based similarity metric." This new approach, which considers perception as probabilistic (CS memory and Perception are shown as distributions), is then fed into a generalization model along with learned values from a learning model (which processes learning data). This contrasts with older approaches that might use more direct or less dynamic perceptual inputs for determining similarity and predicting generalization data. — **Figure 2:** Computational Model Framework Overview. This diagram outlines the architecture of our computational model. It shows how perceptual data is processed by a state-space model (using a Bayesian-like process) to inform a "Model-based similarity metric." This new approach, which considers perception as probabilistic (CS memory and Perception are shown as distributions), is then fed into a generalization model along with learned values from a learning model (which processes learning data). This contrasts with older approaches that might use more direct or less dynamic perceptual inputs for determining similarity and predicting generalization data.

The study demonstrates that understanding this interplay is crucial for our knowledge of generalization behavior and has potential implications for various cognitive domains.

🔗 Read the full scientific publication in ScienceDirect: here

🔍 Your Research Matters — Let’s Share It!

Have you used VSC’s computing power in your research? Did our infrastructure support your simulations, data analysis, or workflow?

We’d love to hear about it!

Take part in our #ShareYourSuccess campaign and show how VSC helped move your research forward. Whether it’s a publication, a project highlight, or a visual from your work, your story can inspire others.

🖥️ Be featured on our website and social media. Show the impact of your work. Help grow our research community

📬 Submit your story: https://www.vscentrum.be/sys