Search for: All records

Physical reservoir computing leverages the intrinsic history-dependence and nonlinearity of hardware to encode spatiotemporal signals directly at the sensor level, enabling low-latency processing of dynamic inputs. Encoding delity depends on the separability of multi-state outputs, yet in practice it is often hampered by empirically chosen, suboptimal operating conditions. Here, we apply Bayesian optimization to improve the encoding performance of solution-processed Al₂O₃/In₂O₃ thin- lm transistors. By exploring a ve-dimensional pulse-parameter input space and using the normalized degree of separation for output state distinguishability, we demonstrate high- delity 6-bit temporal encoding corresponding to 64 output states. We further show that a model based on simpler 4-bit data can effectively guide optimization for more complex 6-bit tasks, substantially reducing experimental effort. Using a six-frame moving-car image sequence as a benchmark, we nd that the optimized 6-bit pulse conditions signi cantly enhance encoding accuracy, with 4-bit derived parameters performing comparably in terms of pixel errors. Shapley Additive Explanations (SHAP) analysis further reveals that gate-pulse amplitude and drain voltage are the dominant contributors to output state separation. This work establishes a data-driven strategy for identifying optimal operating conditions in reservoir devices and outlines a framework that can be transferred to diverse material platforms and physical reservoir implementations. 

Abstract This work identifies and characterizes magnetic structures, especially in terms of small‐scale magnetic flux ropes (SFRs), in the solar wind and magnetosheath across the Earth's bow shock. We investigate the differences between the properties of SFR structures in these regions immediately upstream and downstream of the bow shock by employing two data analysis methods: one based on wavelet transforms and the other based on the Grad‐Shafranov (GS) detection and reconstruction techniques. In situ magnetic field and plasma data from the Magnetospheric Multiscale and Time History of Events and Macroscale Interactions during Substorms missions are used to identify these coherent structures through the two approaches. We identify thousands of SFR event intervals with a range of variable duration over a total time period of 1,000 hr in each region. We report parameters associated with the SFRs such as scale size, duration, magnetic flux content, and magnetic helicity density, derived from primarily the GS‐based analysis results. These parameters are summarized through statistical analysis, and their changes across the bow shock are shown based on comparisons of their respective distributions. We find that in general, the distributions of various parameters follow power laws. The SFR structures seem to be compressed in the magnetosheath, as compared with their counterparts in the solar wind. A significant rotation in the ‐axis defining the orientation of the structures is also seen across the bow shock. We also discuss the implications for the elongation of the SFRs in the magnetosheath along one spatial dimension. 

Abstract Subparsec binary supermassive black holes (BSBHs) should be common from galaxy mergers, yet direct evidence has been elusive. We present Hubble Space Telescope (HST)/WFC3IR F160W imaging for a sample of eight candidate subparsec BSBHs at redshiftsz ∼ 0.1–0.5, as well as cross-comparison with a sample of ordinary quasars with archival HST/WFC3 IR F160W images. These eight candidate subparsec BSBHs were identified from multiepoch spectroscopic surveys of quasars (including both typical quasars and those with single-peaked velocity-offset broad lines), whose broad Hβlines are significantly offset (by ≳ a few hundred kilometers per second) from the systemic redshifts. We directly test the prediction that the host galaxies of BSBHs would have a higher fraction of disturbed morphologies and younger stellar bulges from recent interactions than those of control quasars. After careful subtraction of the central quasar light, our candidate BSBH hosts show a statistically undifferentiated distribution of host asymmetry, indicative of a similar fraction of recent mergers. While a significantly larger sample is needed to place this result on a much firmer statistical ground, it opens questions as to the timescale differences between galaxy merger and BSBH formation, or the efficacy of the radial-velocity-shift-based selection of subparsec BSBH candidates. 

This paper addresses a gap in the AI governance literature in understanding collaboration between national governments and tribal nations in governing AI systems for emergency management. This conceptual work develops and presents a governance design framework for accountable AI systems to fill the knowledge gap by drawing from the fields of public administration, information systems, indigenous studies, and emergency management. This framework situates the governance framework in a cross-sovereignty historical, legal, and policy contexts. It captures the multi-level features and embeddedness of governance structures, including the levels of collaborative governance structure, software system governance rules, and technical software system design. The focal governance dynamics involve the collaborative process in the bi-directional relationship between governance rules and technical design for accountability and the feedback loop. The framework highlights the importance of multi-level and process considerations in designing accountable AI systems. Productive future research avenues include empirical investigation and resulting refinement of the framework and analytical rigor employing institutional grammar. 

Abstract Redox, a native modality in biology involving the flow of electrons, energy, and information, is used for energy‐harvesting, biosynthesis, immune‐defense, and signaling. Because electrons (in contrast to protons) are not soluble in the medium, electron‐flow through the redox modality occurs through redox reactions that are sometimes organized into pathways and networks (e.g., redox interactomes). Redox is also accessible to electrochemistry, which enables electrodes to receive and transmit electrons to exchange energy and information with biology. In this Perspective, efforts to develop electrochemistry as a tool for redox‐based bio‐information processing: to interconvert redox‐based molecular attributes into interpretable electronic signals, are described. Using a series of Case Studies, how the information‐content of the measurements can be enriched using: diffusible mediators; tuned electrical input sequences; and cross‐modal measurements (e.g., electrical plus spectral), is shown. Also, theory‐guided feature engineering approaches to compress the information in the electronic signals into quantitative metrics (i.e., features) that can serve as correlating variables for pattern recognition by data‐driven analysis are described. Finally, how redox provides a modality for electrogenetic actuation is illustrated. It is suggested that electrochemistry's capabilities to provide real‐time, low‐cost, and high‐content data in an electronic format allow the feedback‐control needed for autonomous learning and deployable sensing/actuation.