国家预印本平台

Electrical spin injection and transport in silicon are central challenges for realizing semiconductor-based spintronic devices, particularly in p-type Si, where strong spin relaxation and interface effects often suppress detectable spin signals. Here, we report electrical spin injection, accumulation, and transport in lightly doped p-type silicon using the spin-gapless Heusler compound Mn$_2$CoAl as a ferromagnetic spin injector, separated from the p-Si channel by a thin MgO tunnel barrier in a lateral device geometry. Spin transport is systematically investigated through three-terminal (3-T) Hanle and four-terminal (4-T) nonlocal (NL) spin-valve and Hanle measurements. Clear Lorentzian Hanle signals are observed in the 3-T configuration from 5 K up to room temperature, yielding a spin lifetime of $\sim$0.68 ns at 300 K that increases to $\sim$4.11 ns at 5 K. Temperature-dependent analysis reveals a weak power-law dependence of the spin lifetime, indicating Bir--Aronov--Pikus--type spin relaxation mechanism. To validate genuine spin transport, NL spin-valve and Hanle measurements were performed, revealing well-defined spin-valve switching and controlled spin precession at 5 K. From NL Hanle fitting, a spin lifetime of $\sim$5.65 ns and a spin diffusion length of $\sim$0.82 $μ$m are extracted, confirming diffusive long-range spin transport in the p-Si channel. Although NL signals diminish at elevated temperatures due to reduced interfacial spin polarization and thermal noise, the combined 3-T and 4-T results establish spin-gapless Mn$_2$CoAl as an effective spin injector for p-type silicon. These findings highlight the potential of spin-gapless semiconductors for improving spin injection efficiency and advancing Si-compatible spintronic devices.

Random forests are widely used prediction procedures, yet are typically described algorithmically rather than as statistical designs acting on a fixed set of covariates. We develop a finite-sample, design-based formulation of random forests in which each tree is an explicit randomized conditional regression function. This perspective yields an exact variance identity for the forest predictor that separates finite-aggregation variability from a structural dependence term that persists even under infinite aggregation. We further decompose both single-tree dispersion and inter-tree covariance using the laws of total variance and covariance, isolating two fundamental design mechanisms-reuse of training observations and alignment of data-adaptive partitions. These mechanisms induce a strict covariance floor, demonstrating that predictive variability cannot be eliminated by increasing the number of trees alone. The resulting framework clarifies how resampling, feature-level randomization, and split selection govern resolution, tree variability, and dependence, and establishes random forests as explicit finite-sample statistical designs whose behavior is determined by their underlying randomized construction.

The unprecedented rest-frame UV and optical coverage provided by JWST enables simultaneous constraints on the electron density (n$_{\rm e}$) and temperature (T$_{\rm e}$) of ionized gas in galaxies at z>5. We present a self-consistent direct method based on multiple OIII]1661,66) and [OIII] ($λ$4363, and $λ$5007) transitions to characterize the physical conditions of the high-ionization zone. This new approach is insensitive to a wide range of n$_{\rm e}$ due to the high critical densities of the OIII] and [OIII] transitions. Applying this technique to six galaxies at z=5-9, we find electron densities up to n$_{\rm e}$$\sim 3\times 10^{5}$ cm$^{-3}$ and temperatures of T$_{\rm e}$ $\sim 20,000$ K in systems at $z>6$. Accounting for these self-consistent densities changes the derived T$_{\rm e}$ and modifies the inferred metallicities by up to 0.29 dex relative to previous estimates. We discuss the reported N/O overabundances in the high-$z$ galaxies from our sample, which arise entirely from the high N$^{3+}$/H$^{+}$ values inferred from NIV] lines. We point out that a T$_{\rm e}$-stratification, in which the N$^{3+}$ zone has a slightly higher T$_{\rm e}$ than T$_{\rm e}$([OIII]), could substantially reduce the inferred N/O. Quantitatively, if T$_{\rm e}$(N$^{3+}$) were 10\% higher than T$_{\rm e}$([OIII]), this could induce a systematic overestimation of N$^{3+}$/O$^{2+}$ of nearly 50\%. Classical N/O diagnostics such as N$^{+}$/O$^{+}$, due to their critical densities, can significantly impact the inferred N/O abundance in the presence of high-density gas, whereas N$^{2+}$/O$^{2+}$ place these galaxies closer to $z\sim0$ systems in the N/O-O/H plane. Future JWST programs with larger and more diverse samples will be essential to test the universality and robustness of these results.

Being modeled as a single-label classification task for a long time, recent work has argued that Arabic Dialect Identification (ADI) should be framed as a multi-label classification task. However, ADI remains constrained by the availability of single-label datasets, with no large-scale multi-label resources available for training. By analyzing models trained on single-label ADI data, we show that the main difficulty in repurposing such datasets for Multi-Label Arabic Dialect Identification (MLADI) lies in the selection of negative samples, as many sentences treated as negative could be acceptable in multiple dialects. To address these issues, we construct a multi-label dataset by generating automatic multi-label annotations using GPT-4o and binary dialect acceptability classifiers, with aggregation guided by the Arabic Level of Dialectness (ALDi). Afterward, we train a BERT-based multi-label classifier using curriculum learning strategies aligned with dialectal complexity and label cardinality. On the MLADI leaderboard, our best-performing LAHJATBERT model achieves a macro F1 of 0.69, compared to 0.55 for the strongest previously reported system. Code and data are available at https://mohamedalaa9.github.io/lahjatbert/.

We study the amount of reliable information that can be stored in a DNA-based storage system with noisy sequencing, where each codeword is composed of short DNA molecules. We analyze a concatenated coding scheme, where the outer code is designed to handle the random sampling, while the inner code is designed to handle the random sequencing noise. We assume that the sequencing channel is symmetric and choose the inner coding scheme to be composed by a linear block code and a zero-undetected-error decoder. As a byproduct, the resulting optimal maximum-likelihood decoder land itself for an amenable analysis, and we are able to derive an achievability bound for the scaling of the number of information bits that can be reliably stored. As a result of independent interest, we prove that the average error probability of random linear block codes under zero-undetected-error decoding converges to zero exponentially fast with the block length, as long as its coding rate does not exceed some critical value, which is known to serve as a lower bound to the zero-undetected-error capacity.