Vaikuntanathan Group

We use statistical mechanics, information theory, and machine learning to study how chemical and biological interactions resolve information, from proteins and signaling networks to the cytoskeleton, cell states and neural circuits. We ask the same questions of adaptable synthetic materials, and work on generative AI and statistical mechanics.

Department of Chemistry and the James Franck Institute, University of Chicago

Generating, σ = 40.0

Research

Deciphering the biophysical drivers of interactions

What decides which molecules bind which

1 paper · Read more

CIPHER predicts perturbation responses from the gene–gene covariance of unperturbed cells (Kuznets-Speck et al., Nature Biotechnology, in press).

Cell states, decisions and control

Decision making in high dimensions

4 papers · Read more

Steady-state overlap against memory load α and initial overlap m(0), without (left) and with (right) neuropeptide modulation; the dashed line marks the Hopfield capacity (Goto et al., PNAS, in press).

Chemical drivers of flexible computing in brains

What neuropeptides do for computation in the brain

1 paper · Read more

Cytoskeletal dynamics

Learning and adaptation in cytoskeletal networks

9 papers · Read more

Driven, dissipative excitonic networks as a computational platform (Du et al., arXiv 2026).

Adaptable materials

Memory, learning and computation in synthetic systems

6 papers · Read more

Target, passive-noise and active-noise samples of a Gaussian mixture at three step sizes (Behera et al., Nature Communications, in press).

Generative AI + Stat mech

Connections between modern AI and non-equilibrium physics

6 papers · Read more

Recent News

October 2026: Cal's paper on local imperfect feedback control in non-equilibrium biophysical systems, enabled by response constraints, is accepted at Physical Review X! arXiv
October 2026: CIPHER, our collaboration with the Goyal Lab showing that fluctuation structure predicts genome-wide perturbation outcomes, is accepted at Nature Biotechnology! bioRxiv
September 2026: Daiki and Hector's work with Monika Scholz on neuromodulation-inspired gated associative memory networks is accepted at PNAS! arXiv
September 2026: New on the arXiv: Matt Du shows a coherent advantage in the computational expressivity of excitonic networks, with Cal Floyd and Dipti Jasrasaria. arXiv
2026: Our work showing that non-equilibrium active noise enhances generative memory in diffusion models is accepted at Nature Communications! arXiv
2026: Cal Floyd will join Carnegie Mellon University as an Assistant Professor in January 2027. Congratulations, Cal!
2026: Deb Sankar Banerjee is now an Assistant Professor at the S. N. Bose National Centre for Basic Sciences in Kolkata. Congratulations, Deb!
2026: Jordan Shivers has moved on to a second postdoc at Rice University. Congratulations, Jordan!

Interested in our work? We are always looking for curious minds to join the group. Meet the group · Contact Suri