Publications

Dataset of the paper: "Twisted moiré photonic crystal enabled optical vortex generation through bound states in the continuum".

Localized patterns of amine-terminated monolayers obtained via the surface modification of a monolayer with the biased probe of an atomic force microscope were used to covalently attach buckminsterfullerene or dendrimers to the surface, affording lines as narrow as 20 nm.

We describe observations of the quasi-stellar object (QSO) NRAO 190 during a gamma-ray flare from 1994 August 9 to 1994 August 29. This QSO was serendipitously detected by the EGRET instrument on the Compton Gamma Ray Observatory in a gamma-ray flare with a luminosity at least 10 times that of its quiescent state. Optical, radio, and microwave data were obtained during or near the gamma-ray observations.
\n
\nThe historical behavior of this object places it in the category of bright, flat-spectrum radio sources with strong optical variability that appear to form the largest class of non-Galactic high-energy gamma-ray sources. During the gamma-ray flare the source is observed with l(E > 100 MeV) = 8.4 ± 1.2 × 10-7 photons s-1 cm-2. A single power-law model gives a best-fit photon index of γ = -1.83 ± 0.14.
\n
\nLittle evidence for major radio variability is seen during the flare or immediately afterward, although there is some increase in the 10-100 GHz flux over the next several months. There may be a slight hardening of the radio spectrum. In the optical region there are significant fluctuations on timescales of 1 day or less, although the overall optical luminosity is within the range of previous measurements. Optical observations a few weeks after the gamma-ray observations show a drop of about 60% and reduced variability. A contemporary optical spectrum shows that the source may be slightly harder than seen in a previously published spectrum. Radio monitoring of the source over the year subsequent to the flare has shown a very substantial drop in the flux at many frequencies.

The recent advent of autonomous laboratories, coupled with algorithms for high-throughput screening and active learning, promises to accelerate materials discovery and innovation. As these autonomous systems grow in complexity, the demand for robust and efficient workflow management software becomes increasingly critical. In this paper, we introduce AlabOS, a general-purpose software framework for orchestrating experiments and managing resources, with an emphasis on automated laboratories for materials synthesis and characterization. AlabOS features a reconfigurable experiment workflow model and a resource reservation mechanism, enabling the simultaneous execution of varied workflows composed of modular tasks while eliminating conflicts between tasks. To showcase its capability, we demonstrate the implementation of AlabOS in a prototype autonomous materials laboratory, A-Lab, with around 3,500 samples synthesized over 1.5 years.

The practical usage of reinforcement learning agents is often bottlenecked by\nthe duration of training time. To accelerate training, practitioners often turn\nto distributed reinforcement learning architectures to parallelize and\naccelerate the training process. However, modern methods for scalable\nreinforcement learning (RL) often tradeoff between the throughput of samples\nthat an RL agent can learn from (sample throughput) and the quality of learning\nfrom each sample (sample efficiency). In these scalable RL architectures, as\none increases sample throughput (i.e. increasing parallelization in IMPALA),\nsample efficiency drops significantly. To address this, we propose a new\ndistributed reinforcement learning algorithm, IMPACT. IMPACT extends IMPALA\nwith three changes: a target network for stabilizing the surrogate objective, a\ncircular buffer, and truncated importance sampling. In discrete action-space\nenvironments, we show that IMPACT attains higher reward and, simultaneously,\nachieves up to 30% decrease in training wall-time than that of IMPALA. For\ncontinuous control environments, IMPACT trains faster than existing scalable\nagents while preserving the sample efficiency of synchronous PPO.\n

Prevailing wisdom asserts that one cannot rely on the cloud for critical real-time control systems like self-driving cars. We argue that we can, and must. Following the trends of increasing model sizes, improvements in hardware, and evolving mobile networks, we identify an opportunity to offload parts of time-sensitive and latency-critical compute to the cloud. Doing so requires carefully allocating bandwidth to meet strict latency SLOs, while maximizing benefit to the car.