Research Highlights

In the following paragraphs, I'll give you a glimpse of the research I have been involved in over the last few years at Cornell. For past research, please follow the links on my publications page.

POSTDOCTORAL RESEARCH - How is Performance Evaluation Encoded in the Brain?

 

Dopamine neurons encode performance error in singing birds

 

 

How do organisms associate actions with rewards or punishments? In classic experiments from neuroscience, rats explore an enclosure that has a lever located on one of its walls. When the lever is pressed, a food pellet is dispensed. Soon, through trial and error, the rat learns to press the lever and get the food. In this simple case, the action-reward relationship is clear: press the lever; get the food. But what about in cases in which the reward isn’t so immediately clear, such as learning to speak or play a musical instrument? These skills are not learned for immediate rewards but instead by matching ongoing performance to internal goals.  Neuroscientists have discovered a possible mechanism for simple reward learning, like the rat in the box. In the simple model, neurons from a brain area called the “ventral tegmental area,” or VTA, release the neurotransmitter dopamine in response to a reward. The twist is that these neurons don’t just fire in response to the reward itself, but rather encode “reward prediction error.” This means that the dopamine neurons are activated by better-than-predicted rewards and suppressed by worse-than-predicted rewards. It’s this error signal that appears to drive trial-and-error learning. But whether this error signal exists, and how it might be generated in more complex learning such as speech or music, is unclear.

We study this question in the songbird, which learns to sing by trial and error. “Performance prediction error” in this case is defined by how much the young bird’s song matches the song it has heard from older birds. To compute this error, the songbird must compare its own motor performance (singing) to auditory feedback (how close the song sounds to the ideal version). We record the activity of dopamine neurons in the songbird’s VTA while we distort syllables in the bird’s auditory feedback. Remarkably, we find that these VTA neurons are suppressed immediately after distortions, indicating a worse-than-predicted outcome, and activated at the precise moment in the song when a predicted distortion would have occurred but did not, indicating a better-than-predicted outcome.  This one-dimensional error signal (i.e., better or worse performance) must be computed from auditory evaluation of a complex, high-dimensional motor output (i.e., the song). To investigate how this signal is produced, we will image populations of dopamine neurons and their inputs at cellular resolution. Our work will advance our understanding of learning beyond just a handful of dopamine neurons in the songbird, offering insights into how distributed activity across many brain regions computes the performance prediction error signal fundamental to learning complex skills.

For more information see:

Published Article
PDF

This work has been featured in:
 
Science Magazine Perspective, Simons Foundation Newsletter, CornellCast Video, Cornell Chronicle, The Scientist Magazine, Voice of America, Vice Magazine, Science Daily, Cosmos Magazine, Real Clear Life, Science News, Medium, My Science, Cornell Research, Journal of Experimental Biology

 

GRADUATE RESEARCH - THE SUPERSOLID PROBLEM

 

Introduction (the big picture)

Quantum mechanics is undoubtedly one of the most fascinating aspects of all of physics. The problem with studying quantum phenomena, most of the time however, is that quantum effect are large enough to be directly measurable only when the system size gets extremely small. In most large systems the quantum effects of the constituent atoms and molecules get 'washed out' because the constituent wave functions are out of phase with each other. Is it possible to have a large number of atoms all in phase? Only under very special conditions...

It all began in 1924 when the Indian physicst Satyendra Nath Bose wrote to Albert Einstein describing his ideas on how to treat the electromagnetic waves of the black-body as a gas of identical particles. Einstein immediately saw that Bose was on to someting big and extended Bose's ideas to particles with mass and thus was born Bose-Einstein statistics, the quantum mechanical generalization (for identical particles with integer spin - bosons) of the classical Maxwell-Boltzmann statistics. A profound consequence of Bose-Einstein statistics is the phenomenon of Bose-Einstien Condensation (BEC), which predicts that below a critical temperature, a gas of identical bosons should all (or a macroscopic number of them) settle into the same quantum mechanical ground state. This is an example of a macroscopic quantum phenomenon. We now have a situation where the strange and intriguing effects of quantum mechanics can be observed on a macroscopic (large) scale. This the broad theme of our research.

Classic SupersolidExamples of macroscopic quantum phenomena have been discoverd in gases (Bose-Einstein Condensates) and in liquids (superfluidity in liquid helium-4). It is natural to ask if such phenomena can occur in solids. In superfluid heluim-4, a macroscopic number of atoms get into the ground state and can flow with zero viscosity - hence the term 'superfluid'. It's hard to imagine how such resistance-less flow would occur in a solid. In the late 1960s and early 1970s, it was theoretically proposed by Andreev, Lifshitz, Legett, Reatto and Chester that solid helium-4 might exhibit supersolidity. Their proposal was based on the existance vacancies in the solid helium-4 lattice even at very low temperatures (called zero-point vacancies). The idea was that these vacancies can themselves Bose condense and form a superfluid. As the vacancies move like a superfluid, this is tantamount to the atoms themselves moving with no resistance. It was thus conceived that there could be a lattice of atoms, but a fraction of which could move around with zero viscosity - a supersolid.

For more than three decades, searches for such a state proved fruitless. In 2004, however, Kim and Chan at Penn State found the first evidence that solid helium-4 might display supersolidity. A classical technique for detecting the supefluid transiton uses a torsional oscillator - a container attached to a torsion spring. If the container is filled with liquid helium and the frequency of the oscillator is measured as a function of temperature, one sees an increase in the resonance frequency as the temperature is lowered below the superfluid transition temperature. This is because below the superfluid transition, less and less of the helum is entrained by the wall of the container, thus reducing the effective moment of inertia and increasng the resonance frequency. Interestingly the Penn State group found such an increase in the frequency of a torsion oscillaor even when it was filled with solid helium. This was interpreted as evidence for the elusive supersolid state.

Intense reseach has subsequently been conducted on this phenomenon and the growing evidence points toward far more complex physics than that of the simple supersolid ideas of the 1970s. In fact, it is still an open question as to whether low temperature solid helium is indeed a supersolid. Below are two blurbs describing our groups's efforts (in more technical terms) to unravel the rich mysteries of low temperature solid helium-4.

Interplay of Rotational, Relaxational, and Shear Dynamics in Solid 4He

Interplay FigureWe have developed an ultra-sensitive SQUID-based torsional oscillator (TO) whose motion can be detected either with a SQUID (at low velocities) or capacitively (at high velocities). This device enabled us to map the frequency shift (f) and the dissipation (D) throughout the relevant temperature (T) and velocity (V) range of the proposed supersolid effects (see Figure). To accomplished this comprehensive mapping, we developed the free inertial decay (FID) technique which involves stabiizing T, driving the TO to a high velocity and switching off the drive. As the TO rings down slowly (since it's a very high Q oscillator), the frequency shift and the dissipation can be measured. The fist striking finding from such a map is that the microscopic excitations responsible for the TO effects can be excited both thermally and mechanically, something one would not expect for a simple superfluid. In addition, we find that a measue of the relaxation time of these excitations diverges smoothly and does not exhibit any sharp features. In other words, there is no indication for a thermodynamic critical temperature or velocity. Finally, our approach provides us with an oppotunity to directly compare the TO results with the shear modulus experiments performed by John Beamish and collaborators. We find that the T-V dependence of the TO response is indistinguishable from the T-shear strain dependence of the shear modulus. This means that the excitations responsible for the TO effects mush be the same as those causing the shear effects.

For more information see:

Published Article
PDF

This work has been featured in:
 
Nature News Blog, The Kavli Foundation, Los Alamos News, AAAS EurekAlert, Science Daily, Space Daily, Phys.org

 

Evidence for a Superglass State in Solid 4He

Superglass FigureThe primary finding in this study is the ultra-slow relaxation processes governing the frequency shif (f) and dissipation (D) of the torsion oscillator containing solid helium-4. As the temperature is lowered and we enter the regime of the supersolid effects, we find that the relaxatoin times within f and D begin to increase rapidly. Furthermore, we demonstrate that these relaxation processes in both D and a component of f exhibit a complex synchronized ultra-slow evolution toward equilibrium. To analyse this phenomenon, we introduce the time-dependent Davidson-Cole plot technique (see Figure). The helium dissipation is plotted against the frequency shift of our TO at various temperatures below 300 mK. The resultant surface reveals the delicate origin of glassy defect motions within samples in the proposed phase of supersolidity. We use the generalized rotaional susceptibility approach to provide a theoretical framework in which to analyse these data. We find that while exhibiting these glassy dynamics, the data are quantitatively inconsistent with a simple excitation freeze-out transition, because the variation in f is too large compared to the variation in D (see vertical dashed lines in Figure). This prompts us to propose that if superfluidity is the correct interpretation of the blocked-annulus experiments, then low-temperature solid helium-4 might be desctibed as a superglass.

For more information, see:

Published Article
PDF

This work has been featured in:
 
Science Perspective, Cornell Chronicle, Journal Club for Condensed Matter Physics, Physics World, Physics Today, Questia, Nanowerk