Hyunwoo Gu

This work builds on decades of research on perceptual biases, and it’s been rewarding to contribute to a body of work that first drew me into neuroscience. A joyful journey with my early mentors, Sang-Hun Lee and Sukbin Lim, who continue to challenge me to think across boundaries. (13/13)

Taken together, drift-diffusion dynamics, both shaping and shaped by decision-making, offer a coherent account of behavioral and neural biases. This highlights the value of modeling memory dynamics, beyond static task variables, when explaining complex perceptual behaviors. (12/13)

Trained RNNs suggest possible mechanisms. First, asymmetric feedback from decision to memory populations push memory states toward the chosen directions. Second, a warped representational geometry of orientations induces drift, amplifying decision-consistent bias near diverging stimuli. (11/13)

To explore potential mechanisms, we trained RNNs on the same task structure as human participants. When jointly trained on discrimination and estimation objectives, RNNs developed decision-consistent biases, trading off estimation precision for decision robustness. (10/13)

Next, going beyond sparse behavioral measurements, we tracked working memory states using simultaneously recorded BOLD signals from early visual cortex. Combined decoding and event-based analyses reveal neural signatures consistent with our drift-diffusion scenario. (9/13)

The model makes nuanced predictions about how bias evolves after choice. First, earlier decisions during a delay allow more time for decision-consistent biases to grow. Second, this growth is pronounced near diverging points, and reduced around attractors—consistent with human behavior. (8/13)

The behavioral pattern was consistent with drift-diffusion dynamics. Diffusion can account for the growth of decision-consistent bias, but explaining the growth of stimulus-specific bias requires drift. Model fits further show that participants' behavior is better explained with drift. (7/13)

We first observed previously reported stimulus-specific and decision-consistent biases in participants' behavior. By varying the timing of the discrimination task, our paradigm allowed us to quantify how these biases increase as choice timing is delayed. (6/13)

We tested our model predictions with a paradigm where participants performed both discrimination and estimation tasks on the same stimulus. Long delays between tasks allowed us to track slow memory dynamics through both behavior and fMRI. (5/13)

We modeled these scenarios using a drift-diffusion framework constrained by efficient coding. Drift shifts memory in a stimulus-specific way and can amplify the decision-consistent bias by biasing the choice and then continuing to bias memory in line with that choice. (4/13)

To examine how these biases unfold, we considered two scenarios of underlying memory dynamics. In one, memory becomes noisier through diffusion without systematic direction, consistent with many existing models. In the other, memory also drifts toward stable attractors. (3/13)

We focused on two well-documented types of bias in human vision. The first, stimulus-specific bias, refers to systematic errors tied to objective features of the stimulus. The second, decision-consistent bias, reflects errors aligned with one's own subjective choices. (2/13)

Attractor dynamics of working memory explain a concurrent evolution of stimulus-specific and decision-consistent biases in visual estimation People exhibit biases when perceiving features of the world, shaped by both external stimuli and prior decisions. By tracking behavioral, neural, and mechanistic markers of stimulus- and decision-rela...

Excited to share that our paper is now out in Neuron @cp-neuron.bsky.social (dlvr.it/TM9zJ8).

Our perception isn't a perfect mirror of the world. It's often biased by our expectations and beliefs. How do these biases unfold over time, and what shapes their trajectory? A summary thread. (1/13)