True and false recognition in MINERVA2: Integrating fuzzy-trace theory and computational memory modeling

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Memory Research

Background:

• The MINERVA2 model has previously shown potential in modeling basic Deese/Roediger/McDermott (DRM) false recognition.
• Extending computational models to explain nuanced memory phenomena is crucial for understanding cognitive processes.

Purpose of the Study:

• To enhance the MINERVA2 model by integrating key assumptions from fuzzy-trace theory.
• To demonstrate the model's capability in simulating a broader range of DRM illusion effects and related memory judgments.

Main Methods:

• Incorporated the verbatim-gist distinction using distinct vector representations (distributional semantic for gist, holographic word-form for verbatim).
• Integrated the concept of gist hierarchies by linking item storage quality to semantic similarity with preceding items.
• Tested the enhanced MINERVA2 model against empirical data from DRM illusion studies, including remember/know and source judgments.

Main Results:

• The enhanced MINERVA2 model accurately simulated various empirical effects within the DRM illusion paradigm.
• The model successfully accounted for remember/know and source monitoring judgments, demonstrating its explanatory power.
• The incorporation of gist hierarchies addressed the storage independence problem in multitrace memory models.

Conclusions:

• MINERVA2, augmented with fuzzy-trace theory principles, provides a robust computational framework for modeling the DRM illusion.
• This research strengthens the link between computational modeling approaches and established psychological theories of false memory.
• The findings support MINERVA2 as a scalable tool for investigating episodic memory and false recognition phenomena.