Project | Searching for Semantic Knowledge
I am interested in the storage and retrieval of concepts in semantic memory; your general knowledge of the world and the people, places and objects in it. We can think of semantic memory as a structured space, populated with representations of knowledge. Retrieving a concept from semantic memory is a search process through this space – much like an animal might forage for food in an external environment. We explore the idea of cognitive search using novel Vector Space models and data from patients with focal hippocampal lesions, to create a computational model of restricted semantic search
paper: Searching for Semantic Knowledge: A Vector Space Semantic Analysis of the Feature Generation Task (link to pdf)
TLDR: I sum up the paper in a twitter thread
Project | Inhibition, Competition, and Context Models of Forgetting in Recognition Tests
Causes of forgetting in recognition paradigms investigating Retrieval-Induced Forgetting (RIF) include inhibition (stored traces are reduced in strength due to later practice/testing of related items), competition (increasing trace familiarity adds ‘noise’ when trying to recognize related items), and context change (recognition harmed by context shifts between storage and test). Inhibition predicts decreased recognition for early items from repeated-picture categories compared to non-repeated-picture categories. Competition predicts effects of category size. The results support forgetting caused by competition and familiarity processes, rule out significant effects of inhibition, and by inference suggest that RIF effects seen in prior recognition studies were due to context change.
Project | Neural Mechanisms of Category Integration
We are looking to characterize the neuroanatomical substrates of category information, and how this relates to mechanisms of memory search in both episodic and semantic domains. The idea of a gradually changing contextual representation is central to retrieved-context models of human memory. Items are recalled closer together in a recall sequence if they share similar features; whether they are semantically related, temporally adjacent, or clustered by category. We use neuroimaging methods (fMRI & EEG) to elucidate a neural signature of this integrative mechanism.
Current opinions paper: Retrieved-context models of memory search and the neural representation of time (link to pdf)
CNS2020 poster: Characterizing the interaction of temporal and semantic
information in categorized memory search (link to pdf)
Project | Interaction of Episodic/Semantic
Our experience of time is continuous and ongoing as it happens, but when we look back along our mental timeline it is packaged into discrete events. We take an integrative approach to the study of memory. to understand how semantic information might interact with temporal information in a way that makes some events more memorable than others. We use computational models to investigate what is (dis) similar in these memory systems, and the underlying mechanisms by which memory for events and knowledge is encoded, maintained, and retrieved.
Collaborators:
Audrey Duarte Ph.D (Mentor) · Memory & Aging Lab | UT Austin
Sean Polyn Ph.D (Advisor) · Computational Memory Lab | Vanderbilt
Ashleigh Maxcey Ph.D · Maxcey Memory Lab | Vanderbilt
Richard Shiffrin Ph.D · Memory and Perception Lab | Indiana
Robert Nosofsky Ph.D · Categorization and Memory Lab | Indiana
Melissa Duff Ph.D · Communication and Memory | Vanderbilt
Sarah Brown-Schmidt Ph.D · Conversation Lab | Vanderbilt
Neal Morton Ph.D · Center for Learning and Memory | UT Austin
Rebecca Cutler 