Publications

Structural investigations of bare and surface-modified polystyrene microspheres (beads) have been carried out by infrared-visible sum frequency generation (SFG) vibrational spectroscopy and scanning force microscopy (SFM). Bead surfaces have been modified by either the covalent linking of immunoglobulin G (IgG) and bovine serum albumin (BSA) or the nonspecific adsorption of a Pluronic surfactant. After surface modification with protein, SFG signals in the aliphatic CH-stretch region are detected at both the buffer/bead and air/bead interfaces, indicating that some amino acid residues in proteins adopt preferred orientations. SFG results indicate that the hydrophobic poly(propylene glycol) moieties in the Pluronic order when adsorbed onto the bead, at both the buffer/bead and air/bead interfaces, whereas hydrophilic poly(ethylene glycol) groups align to a lesser extent. SFG spectra also show that the phenyl rings of bare polystyrene beads in contact with air or buffer are ordered, with a dipole component directed along the surface normal, but become less ordered after the adsorption of either proteins or the polymer. Molecular orientation and ordering at the bead surface affect its hydrophobicity and aggregation behavior. SFM results reveal the formation of nonuniform islands when bare beads with more hydrophobic character are spun-cast onto a silica substrate. In the presence of adsorbed protein, a hexagonal packing of beads, with some defects, is observed, depending on the bulk pH and the type of attached protein. Adsorbed Pluronic causes the beads to aggregate in a disordered fashion, as compared to the behavior of bare and protein-modified beads.

Treatment with acetyl-CoA carboxylase inhibitors (ACCi) in nonalcoholic steatohepatitis (NASH) may increase plasma triglycerides (TGs), with variable changes in apoB concentrations. ACC is rate limiting in de novo lipogenesis and regulates fatty acid oxidation, making it an attractive therapeutic target in NASH. Our objectives were to determine the effects of the ACCi, firsocostat, on production rates of plasma LDL-apoB in NASH and the effects of combined therapy with fenofibrate. Metabolic labeling with heavy water and tandem mass spectrometric analysis of LDL-apoB enrichments was performed in 16 NASH patients treated with firsocostat for 12 weeks and in 29 NASH subjects treated with firsocostat and fenofibrate for 12 weeks. In NASH on firsocostat, plasma TG increased significantly by 17% from baseline to week 12 (P = 0.0056). Significant increases were also observed in LDL-apoB fractional replacement rate (baseline to week 12: 31 ± 20.2 to 46 ± 22.6%/day, P = 0.03) and absolute synthesis rate (ASR) (30.4-45.2 mg/dl/day, P = 0.016) but not plasma apoB concentrations. The effect of firsocostat on LDL-apoB ASR was restricted to patients with cirrhosis (21.0 ± 9.6 at baseline and 44.2 ± 17 mg/dl/day at week 12, P = 0.002, N = 8); noncirrhotic patients did not change (39.8 ± 20.8 and 46.3 ± 14.8 mg/dl/day, respectively, P = 0.51, N = 8). Combination treatment with fenofibrate and firsocostat prevented increases in plasma TG, LDL-apoB fractional replacement rate, and ASR. In summary, in NASH with cirrhosis, ACCi treatment increases LDL-apoB100 production rate and this effect can be prevented by concurrent fenofibrate therapy.

Large language models (LLMs) have rapidly improved text embeddings for a growing array of natural-language processing tasks. However, their opaqueness and proliferation into scientific domains such as neuroscience have created a growing need for interpretability. Here, we ask whether we can obtain interpretable embeddings through LLM prompting. We introduce question-answering embeddings (QA-Emb), embeddings where each feature represents an answer to a yes/no question asked to an LLM. Training QA-Emb reduces to selecting a set of underlying questions rather than learning model weights. We use QA-Emb to flexibly generate interpretable models for predicting fMRI voxel responses to language stimuli. QA-Emb significantly outperforms an established interpretable baseline, and does so while requiring very few questions. This paves the way towards building flexible feature spaces that can concretize and evaluate our understanding of semantic brain representations. We additionally find that QA-Emb can be effectively approximated with an efficient model, and we explore broader applications in simple NLP tasks.

A longstanding performance challenge in datacenter-based applications is how to efficiently handle incoming client requests that spawn many very short (μs scale) jobs that must be handled with high throughput and low tail latency. When no assumptions are made about the duration of individual jobs, or even about the distribution of their durations, this requires blind scheduling with frequent and efficient preemption, which is not scalably supported for μs-level tasks. We present Tiny Quanta (TQ), a system that enables efficient blind scheduling of μs-level workloads. TQ performs fine-grained preemptive scheduling and does so with high performance via a novel combination of two mechanisms: forced multitasking and two-level scheduling. Evaluations with a wide variety of μs-level workloads show that TQ achieves low tail latency while sustaining 1.2x to 6.8x the throughput of prior blind scheduling systems.

Abstract We present the discovery and analysis of SN 2022oqm, a Type Ic supernova (SN) detected <1 day after the explosion. The SN rises to a blue and short-lived (2 days) initial peak. Early-time spectral observations of SN 2022oqm show a hot (40,000 K) continuum with high ionization C and O absorption features at velocities of 4000 km s −1 , while its photospheric radius expands at 20,000 km s −1 , indicating a pre-existing distribution of expanding C/O material. After ∼2.5 days, both the spectrum and light curves evolve into those of a typical SN Ic, with line velocities of ∼10,000 km s −1 , in agreement with the evolution of the photospheric radius. The optical light curves reach a second peak at t ≈ 15 days. By t = 60 days, the spectrum of SN 2022oqm becomes nearly nebular, displaying strong Ca ii and [Ca ii ] emission with no detectable [O i ], marking this event as Ca-rich. The early behavior can be explained by 10 −3 M ⊙ of optically thin circumstellar material (CSM) surrounding either (1) a massive compact progenitor such as a Wolf–Rayet star, (2) a massive stripped progenitor with an extended envelope, or (3) a binary system with a white dwarf. We propose that the early-time light curve is powered by both the interaction of the ejecta with the optically thin CSM and shock cooling (in the massive star scenario). The observations can be explained by CSM that is optically thick to X-ray photons, is optically thick in the lines as seen in the spectra, and is optically thin to visible-light continuum photons that come either from downscattered X-rays or from the shock-heated ejecta. Calculations show that this scenario is self-consistent.

Abstract An integrated geostatistical methodology for a site-specific reservoir description was used in reservoir simulations to establish additional recovery benefits for gravity drainage infill wells for the Romeo interval for Drill Sites 2 and 5 at Prudhoe Bay. The available geological information varied over multiple scales and included: a detailed chrono-stratigraphic framework relating rocks of similar depositional times, a detailed depositional model with facies associations described at each well from log and core data, and analog outcrop training images obtained using photomosaics. The reservoir description methodology transformed the facies associations at each well into indicator variables and used these variables to analyze vertical and lateral spatial correlation. Conditional truncated gaussian simulation was then used to generate 3D facies association maps within the model area constrained by the chrono-stratigraphic framework. Input from geologists was critical in the evaluation of the geological realism of the computed facies association maps. Porosity and permeability vertical and lateral spatial correlations were then determined for each facies association. Vertical correlation was determined from variogram analysis while lateral correlation was determined from the outcrop training images. Conditional sequential gaussian simulation was then used to distribute 3D porosity and permeability, separately within each facies association, throughout the model area. A pressure solver was used to scale up the 3D fine scale permeability distributions to grid blocks appropriate for direct input into a flow simulation model. The conditional simulation methods allowed the integration of geological information on a number of different scales, thereby capturing the 3D complexity of the Romeo interval. The 3D facies association description generated using this methodology was judged by geologists to depict a very reasonable geological picture of the DS-2,5 Romeo interval. The pressure solver used to scale up the fine scale permeability distributions maintained the major permeability features apparent in the fine scale distribution which were important to reservoir performance.

Corrosion resistance of titanium (Ti) alloys is closely connected with their microstructure which can be adjusted and controlled via different annealing schemes. Herein, we systematically investigate the specific effects of annealing on the corrosion performance of Ti-6Al-3Nb-2Zr-1Mo (Ti80) alloy in 3.5 wt.% NaCl and 5 M HCl solutions, respectively, based on open circuit potential (OCP), potentiodynamic polarization, electrochemical impedance spectroscopy (EIS), static immersion tests and surface analysis. Results indicate that increasing annealing temperature endows Ti80 alloy with a higher volume fraction of β phase and finer α phase, which in turn improves its corrosion resistance. Surface characterization demonstrates that β phase is more resistant to corrosion than α phase owing to a higher content of Nb, Mo, and Zr in the former; additionally, the decreased thickness of α phase alleviates segregation of elements to further restrain the micro-galvanic couple effects between α and β phases. Meanwhile, the influential mechanisms of environmental conditions on corrosion of Ti80 alloy are discussed in detail. As the formation of a highly compact and stable oxide film on surface, annealed Ti80 alloys exhibit a low corrosion current density (10−6 A/cm2) and high polarization impedance (106 Ω⋅ cm2) in 3.5 wt.% NaCl solution. However, they suffer severe corrosion in 5 M HCl solution, resulting from the breakdown of native oxide films (the conversion of TiO2 to aqueous Ti3+), active dissolution of substrate Ti to aqueous Ti3+ and existence of micro-galvanic couple effects. Those findings could provide new insights to designing Ti alloys with high-corrosion resistance through microstructural optimization.

Extensive early observations proved that the ejecta of supernova 1987A (SN 1987A) are aspherical. The most important of these early observations include (1) the "Bochum event" that revealed small-scale spectroscopic structure indicating chemical inhomogeneities of the ejecta, (2) spectropolarimetry that showed deviations from spherical symmetry, and (3) speckle observations that revealed both the asymmetry of the ejecta and the "mystery spot" manifested as a secondary source off center from the bulk of the supernova ejecta. Fifteen years after the supernova explosion, the Hubble Space Telescope has resolved the rapidly expanding ejecta. The late-time images and spectroscopy provide a geometrical picture that is consistent with early observations and suggests a highly structured, axially symmetric geometry. We present here a new synthesis of the old and new data. We show that the Bochum event, presumably a clump of 56Ni, and the late-time image, the locus of excitation by 44Ti, are most naturally accounted for by sharing a common position angle of about 14°, the same as the mystery spot and early speckle data on the ejecta, and that they are both oriented along the axis of the inner circumstellar ring at 45° to the plane of the sky. We also demonstrate that the polarization represents a prolate geometry with the same position angle and axis as the early speckle data and the late-time image, and hence that the geometry has been fixed in time and throughout the ejecta. The Bochum event and the Doppler kinematics of the [Ca II]/[O -II] emission in spatially resolved Hubble Space Telescope spectra of the ejecta can be consistently integrated into this geometry. The radioactive clump is deduced to fall approximately along the axis of the inner circumstellar ring and therefore to be redshifted in the north, whereas the [Ca II]/[O II] 7300 Å emission is redshifted in the south. We present a jet-induced model for the explosion and argue that such a model can account for many of the observed asymmetries. In the jet models, the oxygen and calcium are not expected to be distributed along the jet but primarily in an expanding torus that shares the plane and northern blue shift of the inner circumstellar ring.