Memory "in the wild"

Memory has been typically studied in the laboratory using random lists of words stripped of the rich contextual structures that usually accompany our normal everyday experiences. Neisser (1976) argued that cognitive psychology had failed to address everyday human behavior due to its overreliance on artificial laboratory-based experiments. While it was not clear at that time how one would develop an ecologically valid memory science, today we are able to rigorously capture human experience using wearable devices and smartphones.

We use lifelogging technologies to capture and quantify human experience as it happens in the real world and use this information to design memory, learning, and decision-making tasks which are administered both in the laboratory and during the course of people's everyday experience, with or without concurrent brain recordings (EEG/iEEG/fMRI). Complementing these naturalistic memory studies, we also perform more controlled laboratory studies using virtual reality with a focus on understanding the role of different dimensions of context in memory. Using such a multi-pronged approach, we will provide critical real-world tests of theories of memory and event cognition that were developed in highly contrived laboratory studies.

Please check out the publications page for a list of prior work on related topics (e.g. Sreekumar et al., 2014; Nielson et al., 2015; Sreekumar et al., 2018).

The lab seeks students with strong quantitative and coding skills and an interest in human memory to work on projects in this area.


Inspired by Walter J Freeman's pioneering work in neurodynamics, we seek to understand how different propagating patterns of neural activity play a role in cognition. We use methods from the physics of wave propagation and apply stability analysis of nonlinear dynamical systems to characterize complex and plane propagating waves in the brain. Shared invasive brain recordings from humans and animals will be used in this endeavor as well as scalp EEG recordings in new experiments designed to test hypotheses about the functional role of propagating waves. See Sreekumar et al. (2020) on the publications page for a characterization of propagating plane waves of neural activity and their relationship with traveling waves at the micro-scale as well as single unit spiking in the human brain.

We are also interested in other types of neural dynamics such as an internally drifting temporal context (Howard & Kahana, 2002) and their role in memory and learning. See for example El-Kalliny et al. (2019) on the publications page.

The lab seeks students with strong quantitative and coding skills and expertise/interest in the physics of complex systems to work on projects in this area.

Mental health applications

We will build tools that are rooted deeply in the basic research that we and others do in order to help society at large deal with cognitive and mental health issues. An important aspect of this endeavor would be to collect lifelogging data at a large scale and over long time periods within the Indian population, and also from key clinical subpopulations of interest such as mild cognitive impairment and Alzheimer's patients. We will use dynamical systems methods to quantify individuals' experience and apply causal inference techniques to understand how variables tracked by different sensors influence cognitive and mental outcomes. We will also build mobile apps gamifying certain types of cognitive training that have been found to be beneficial for improving performance in different cognitive domains but by using individually tailored stimuli obtained from people's own lifelogs.

The lab seeks students with expertise/interest in the physics of dynamical systems, AI, and machine learning to work on projects in this area.