Physics Colloquia

How does learning occur? Neural networks learn via optimization, where a loss function is minimized by a computer to achieve the desired result. But physical networks such as mechanical spring networks or flow networks have no central processor so they cannot minimize such a loss function. An alternative is to encode local rules into those networks so that they can evolve under external driving to develop function. For example, if the springs in a mechanical network have equilibrium lengths that grow if the springs are stretched, and shrink when the springs are compressed, the network will naturally evolve under applied stresses. I will describe how both of these strategies—global minimization of a loss function as well as training by local rules--can be used to teach systems how to perform functions inspired by biology, such as the ability of proteins (e.g. hemoglobin) to change their conformations upon binding of an atom (oxygen) or molecule, the ability of the brain’s vascular network to send enhanced blood flow and oxygen to specific areas of the brain associated with a given task, or the ability of the brain to perform classification.

Topology plays a key role in condensed matter physics, underlying much of our understanding of equilibrium matter in terms of defects in ordered media and topologically protected states. In active systems – collections of entities that consume energy to generate their own motion and forces – topological phenomena can take on new and surprising roles. I will describe some of these behaviors focusing on liquid-crystalline active matter in two dimensions, specifically active nematic liquid crystals, where defects become motile particles, drive spatio-temporally chaotic flows, and can themselves organize in emergent ordered states.

There are many intriguing physical phenomena that are associated with topological features --- global properties that are not discernible locally. The best-known examples are quantum Hall effects in electronic systems, where insensitivity to local properties manifests itself as conductance through edge states which are insensitive to defects and disorder. In the talk, we first discuss how similar physics can be explored with photons; specifically, how various topological models can be simulated in various photonics systems, from ring resonators to photonic crystals. We then discuss that the integration of strong optical nonlinearity can lead to unique bosonic phenomena, such as topological frequency combs, topological source of quantum light and chiral quantum optics. These results may enable the development of classical and quantum optical devices with built-in protection for next-generation optoelectronic and quantum technologies. In the end, we discuss an emerging field at the interface of quantum optics and correlated electron systems, with the goal of creating and manipulating many-body states of light-matter hybrids with new functionalities, such as high-Tc superconductors.