Clinician-EngineerResearch Engineer

Ian Taylor

Working where psychology meets machine learning. I decode visual experience from brain signals - EEG, representational similarity analysis, and multivariate pattern analysis - to understand how minds, and models, represent the world.

01

Background

MS, Computer Science (AI concentration)

Kennesaw State UniversityKennesaw, GAGraduated May 15, 2026

Thesis: “EEG Decoding and Representational Similarity Analysis of Visual Scenes,” advised by Dr. Selena He and Dr. Tim Martin. Focus on RSA, EEG signal processing, and MVPA.

Psychometrist

Georgia Psychological AssociatesMarietta, GAClinical practice

Administered and scored 2,000+ standardized neuropsychological assessments - WAIS-IV, WISC-IV/V, and KBIT - building a clinician's intuition for how cognition is measured.

2,000+

Neuropsych assessments scored

3

Test batteries (WAIS · WISC · KBIT)

1

MS thesis on EEG visual decoding

02

Research interests

NeuroAI

Drawing the brain and artificial networks into a shared representational frame.

Brain-computer interfaces

Decoding intent and percept from non-invasive neural signals.

Neuroprosthetics

Closing the loop between neural decoding and assistive control.

Computational neuroscience

Modeling how populations of neurons encode the visual world.

Visual imagery & aphantasia

Why some minds picture and others don’t - and what that reveals about cognition.

03

Selected writing

Co-authorMay 2026

Evolutionary Pressures on Mental Imagery, External Symbolic Storage, and Adaptive Accounts of Aphantasia

With Merlin Monzel. An adaptive account of why mental imagery varies across individuals - and how external symbolic storage may relax the selective pressure to picture things internally.

04

Live paper feed

Recent preprints across cognitive science, computational neuroscience, and AI/ML - the literature I read to stay current. Filter by domain or search across abstracts.

Live from arXiv (cs.AI, cs.LG, q-bio.NC) and bioRxiv (neuroscience), sorted by submission date. Parsed client-side from the public APIs.