Explore the vast range of titles available.

Turbulent magnetic fields abound in nature, pervading astrophysical, solar, terrestrial and laboratory plasmas. Understanding the ubiquity of magnetic turbulence and its role in the universe is an outstanding scientific challenge. Here, we report on the transition of magnetic turbulence from an initially electron-driven regime to one dominated by ion-magnetization in a laboratory plasma produced by an intense, table-top laser. Our observations at the magnetized ion scale of the saturated turbulent spectrum bear a striking resemblance with spacecraft measurements of the solar wind magnetic-field spectrum, including the emergence of a spectral kink. Despite originating from diverse energy injection sources (namely, electrons in the laboratory experiment and ion free-energy sources in the solar wind), the turbulent spectra exhibit remarkable parallels. This demonstrates the independence of turbulent spectral properties from the driving source of the turbulence and highlights the potential of small-scale, table-top laboratory experiments for investigating turbulence in astrophysical environments. Understanding the role of magnetic turbulence in the atmosphere is difficult as direct access is limited, but latest laser technology can enable such studies in the lab. Here the authors probe the evolution of such turbulence in laser-generated plasma with its implications to astrophysical environments.

Shock waves in plasmas can be characterized by the mechanisms behind their formation. When binary collisions are frequent, dissipation is collision-driven and the shock width is a few mean free paths. In contrast, collisionless shocks rely on collective plasma processes to establish dissipation on scales well below the mean free path. We bridge these regimes with particle-in-cell simulations using OSIRIS with a Coulomb-collision module, varying parameters that control collisionality. We find a smooth transition of the shock width in the intermediate region where the ion plasma parameter \\(N_D 1\\). Our results recover the asymptotic predictions: a collisional-regime width consistent with the Mott-Smith ansatz with a BGK operator, and the collisionless limit consistent with Tidman's classical formalism. The ion plasma parameter thus serves as a practical metric for identifying when shocks shift from fluid-like, mean-free-path scales to collisionless, sub-mean-free-path scales. We discuss implications for astrophysical environments, where shock breakout changes the shock width and marks the onset of efficient particle acceleration.

We propose and simulate a laboratory platform to study the effects of positrons in magnetic reconnection using laser-driven capacitor coils. Using particle-in-cell simulations, we show that externally injected MeV electron-positron pairs are trapped in the coil current sheet, significantly modifying the reconnection dynamics and particle acceleration. These pairs increase the reconnection rate by a factor of approximately 8, which Ohm's law decomposition reveals to be driven by the divergence of the generalized pressure tensor. Based on their high energy and magnetization, the pairs also substantially broaden the diffusion region. Particle tracking simulations in realistic coil magnetic fields further demonstrate that injected pairs can remain confined for several picoseconds, providing conditions for sustained interaction with the reconnection region. These results establish a near-term pathway to laboratory studies of positron-influenced reconnection, bridging high-energy-density experiments with pair-dominated astrophysical environments.

Magnetic field generation in ultra-intense laser-solid interactions is studied over a range of laser intensities relevant to next-generation laser facilities (\\(a_0 = 50-500\\)) using 2D particle-in-cell simulations. It is found that fields on the order of 0.1 MT (1 GigaGauss) may be generated by relativistic electrons traveling along the surface of the target. However a significant fraction of the energy budget is converted to high-energy photons, ~38% at \\(a_0=500\\), greatly reducing the available energy for field generation. A model for the evolution of the target-surface fields and their scaling with \\(a_0\\) is developed using laser parameters and assumed values for the average radial electron velocity and reflectivity. The model and empirical scaling allow for the estimation of field strengths on the next generation of laser facilities, a necessary component to the proposal of any future magnetized experiment.

Large language models (LLMs) and medical AI tools are routinely used by physicians and patients for medical advice, yet their clinical safety profiles remain poorly characterized. We present NOHARM (Numerous Options Harm Assessment for Risk in Medicine), a 1,100-task benchmark of primary care-to-specialist consultation cases to measure the frequency and severity of potentially harmful errors from LLM-generated medical consultation recommendations. NOHARM covers 10 specialties, with 12,747 expert annotations for 4,249 clinical management options. Across 20 notable LLMs and 4 widely used retrieval-augmented generation (RAG) clinical AI tools, direct application of recommendations carried potential for severe harm in up to 24.6% of cases, with errors of omission accounting for more than 80% of severe errors. Harm potential was not uniform across systems, with clinical AI tools outperforming generalist LLMs, and multi-agent AI teaming further improving performance in generalist models. In a randomized study of 101 U.S.-licensed generalist physicians, AI assistance improved physician performance compared to conventional resources. However, AI-assisted physicians frequently omitted valuable AI-generated recommendations and still scored lower than many AI systems alone. Had those recommendations been incorporated, combined human-AI responses would have outperformed both the human and AI system as used, suggesting complementary strengths and unrealized potential in human-AI teaming. Collectively, these results show that despite strong performance on medical knowledge benchmarks, widely used AI tools can produce medical consultation advice with the potential for severe harm, and highlight the need for explicit measurement of clinical safety. The benchmark and leaderboard are publicly available to support ongoing evaluation and improvement of AI systems used for clinical care.