Search for: All records

Circular RNAs (∼16−44 nt) were enzymatically synthesized efficientlyviaa novel DNA dumbbell splinting strategy, further, the circular 44 nt RNA was used as scaffold strands to construct hybrid and pure RNA double crossover tiles and nanostructures. 

Kernel-based learning algorithms have been extensively studied over the past two decades for their successful applications in scientific research and industrial problem-solving. In classical kernel methods, such as kernel ridge regression and support vector machines, an unregularized offset term naturally appears. While its importance can be defended in some situations, it is arguable in others. However, it is commonly agreed that the offset term introduces essential challenges to the optimization and theoretical analysis of the algorithms. In this paper, we demonstrate that Kernel Ridge Regression (KRR) with an offset is closely connected to regularization schemes involving centered reproducing kernels. With the aid of this connection and the theory of centered reproducing kernels, we will establish generalization error bounds for KRR with an offset. These bounds indicate that the algorithm can achieve minimax optimal rates. 

One of the challenges for multiagent reinforcement learning (MARL) is designing efficient learning algorithms for a large system in which each agent has only limited or partial information of the entire system. Whereas exciting progress has been made to analyze decentralized MARL with the network of agents for social networks and team video games, little is known theoretically for decentralized MARL with the network of states for modeling self-driving vehicles, ride-sharing, and data and traffic routing. This paper proposes a framework of localized training and decentralized execution to study MARL with the network of states. Localized training means that agents only need to collect local information in their neighboring states during the training phase; decentralized execution implies that agents can execute afterward the learned decentralized policies, which depend only on agents’ current states. The theoretical analysis consists of three key components: the first is the reformulation of the MARL system as a networked Markov decision process with teams of agents, enabling updating the associated team Q-function in a localized fashion; the second is the Bellman equation for the value function and the appropriate Q-function on the probability measure space; and the third is the exponential decay property of the team Q-function, facilitating its approximation with efficient sample efficiency and controllable error. The theoretical analysis paves the way for a new algorithm LTDE-Neural-AC, in which the actor–critic approach with overparameterized neural networks is proposed. The convergence and sample complexity are established and shown to be scalable with respect to the sizes of both agents and states. To the best of our knowledge, this is the first neural network–based MARL algorithm with network structure and provable convergence guarantee. 

Nitrogen doped lutetium hydride has drawn global attention in the pursuit of room-temperature superconductivity near ambient pressure and temperature. However, variable synthesis techniques and uncertainty surrounding nitrogen concentration have contributed to extensive debate within the scientific community about this material and its properties. We used a solid-state approach to synthesize nitrogen doped lutetium hydride at high pressure and temperature (HPT) and analyzed the residual starting materials to determine its nitrogen content. High temperature oxide melt solution calorimetry determined the formation enthalpy of LuH_1.96N_0.02(LHN) from LuH₂and LuN to be −28.4 ± 11.4 kJ/mol. Magnetic measurements indicated diamagnetism which increased with nitrogen content. Ambient pressure conductivity measurements observed metallic behavior from 5 to 350 K, and the constant and parabolic magnetoresistance changed with increasing temperature. High pressure conductivity measurements revealed that LHN does not exhibit superconductivity up to 26.6 GPa. We compressed LHN in a diamond anvil cell to 13.7 GPa and measured the Raman signal at each step, with no evidence of any phase transition. Despite the absence of superconductivity, a color change from blue to purple to red was observed with increasing pressure. Thus, our findings confirm the thermodynamic stability of LHN, do not support superconductivity, and provide insights into the origins of its diamagnetism. 

Abstract This study presents new experimental data on the thermodynamic stability of SiC(O) and SCN(O) ceramics derived from the pyrolysis of polymeric precursors: SMP‐10 (polycarbosilane), PSZ‐20 (polysilazane), and Durazane‐1800 (polysilazane) at 1200°C. There are close similarities in the structure of the polysilazanes, but they differ in crosslinking temperature. High‐resolution X‐ray photoelectron spectroscopy shows notable differences in the microstructure of all polymer‐derived ceramics (PDCs). The enthalpies of formation (∆H°_{f, elem}) of SiC(O) (from SMP‐10), SCN(O) (from PSZ‐20), and SCN(O) (from Durazane‐1800) are −20 ± 4.63, −78.55 ± 2.32, and −85.09 ± 2.18 kJ/mol, respectively. The PDC derived from Durazane‐1800 displays greatest thermodynamic stability. The results point to increased thermodynamic stabilization with addition of nitrogen to the microstructure of PDCs. Thermodynamic analysis suggests increased thermodynamic drive for forming SiCN(O) microstructures with an increase in the relative amount of SiN_xC_4−xmixed bonds and a decrease in silica. Overall, enthalpies of formation suggest superior stabilizing effect of SiN_xC_4−xcompared to SiO_xC_4−xmixed bonds. The results indicate systematic stabilization of SiCN(O) structures with decrease in silicon and oxygen content. The destabilization of PDCs resulting from higher silicon content may reach a plateau at higher concentrations.