Cate Hall

Einstein Probe Detects First X-ray Shock Breakout from Type Ic-BL Supernova, Revealing Diverse SBO Luminosity

The Einstein Probe (EP) has detected EP260321a, an FXT with X-ray shock breakout (SBO) emission, associated with the Type Ic-BL supernova SN 2026gzf. Multi-wavelength follow-up, including radio observations, constrains the ejecta properties and rules out high-energy jets. The light curve is best explained by interaction with circumstellar material and radioactive decay. While SN 2026gzf aligns with typical Type Ic-BL SN properties, the inferred rate of similar X-ray SBO events suggests that many such supernovae may produce weaker X-ray signals than observed for EP260321a.

WildCross Benchmark Enhances Robotic Perception in Natural Environments

The WildCross benchmark addresses the limitations of current robotic perception systems, particularly vision foundation models, in natural environments. It provides a large-scale dataset with synchronized RGB, depth, surface normal, Lidar, and pose data for evaluating place recognition and metric depth estimation. This work offers an expanded analysis of the benchmark, with a focus on depth estimation.

Depth-Aware Distillation Enhances Visual Place Recognition in Forests

This paper introduces a lightweight, depth-aware distillation framework to enhance visual place recognition in challenging forest environments. By integrating geometric cues from depth data into a DINOv2-based model without altering its descriptor space, the approach improves robustness against appearance variations. Evaluations on the WildCross benchmark confirm the effectiveness of depth as a complementary modality, positioning depth-aware distillation as a key technique for more reliable forest perception systems.

Reliability Assessment and Novel Selection Method for AGN Properties in Euclid Q1 Data

The Euclid Quick Data Release (Q1) enables the study of Active Galactic Nuclei (AGN) and their host galaxies. A novel method employing mock Spectral Energy Distributions (SEDs) of AGN and galaxies is introduced to assess the reliability of derived physical properties like stellar mass and AGN luminosity. This method, applied to the Q1 catalog, yields high reliability for a significant portion of sources and introduces a new AGN selection criterion based on the AGN fraction of mid-infrared flux, which demonstrates comparable performance to traditional color-color selections but with improved redshift independence and sensitivity to fainter objects. The study underscores the necessity of combining multiple AGN identification techniques to mitigate selection biases and achieve a more comprehensive understanding of AGN and galaxy evolution.

MAFS: A Parameter-Economical Alternative to Froggatt-Nielsen for Fermion Mass Hierarchies

The authors introduce a framework called Maximal Abelian Flavor Symmetries (MAFS) that explains the observed hierarchies in quark and lepton masses and mixing angles using a minimal set of small parameters—one per fermion multiplet. Unlike Froggatt-Nielsen, MAFS has no free charge assignments, making it universally applicable and increasingly predictive as the number of multiplets decreases. In SU(5) GUTs, 15 observed mass ratios and mixing angles are described at the factor-of-two level by just five parameters; in SO(10), only three parameters suffice. The framework further predicts that the observed quark hierarchical pattern necessitates large neutrino mixing angles and small neutrino mass hierarchies, and that the baryon asymmetry can arise from leptogenesis without additional small parameters in SU(5).

Broad Temporal Sampling Beats Temporal Continuity for ML Climate Downscaling Under Fixed Data Budgets

When allocating limited high-resolution climate simulation years for training ML downscalers, distributing training years across the full climate trajectory outperforms both historical-only and endpoint-only sampling. The key insight is that broad coverage of internal climate variability provides more marginal value than temporal continuity or simply capturing the forced response from past and future bookends. Even using just 10% of available years, temporally distributed training matches near-full-data performance, directly informing how to allocate scarce computational resources for regional climate emulation.

Multidimensional Transparency Framework for Multi-Agent LLM Systems Identified in Early Adopter Conceptualizations

A study of early adopters building multi-agent LLM systems reveals a nuanced understanding of transparency, encompassing reproducibility, debugging, boundary-setting, visualization, and auditing. These diverse perspectives highlight transparency as a situated, socio-technical practice essential for responsible AI development in complex, distributed systems. The findings inform future HCI and AI research by emphasizing the need to align transparency mechanisms with the expectations and capabilities of all stakeholders, including developers, users, and governance bodies.

Argon seeding enables GW-class fusion power with divertor detachment in ARC

Integrated core-edge modeling of the ARC tokamak shows that Argon seeding can simultaneously achieve fusion power approaching 1 GW and divertor temperatures below 2 eV, satisfying detachment requirements. Performance is strongly sensitive to separatrix density (750-1000 MW range) but only weakly dependent on enrichment factor and pedestal density. Neon seeding yields lower performance (600-850 MW) and worse H-mode access due to core impurity accumulation, while tungsten impurity peaking remains modest and decreases at higher Zeff.

Accelerating Surface Radiation Content to Investigate the Impact of Radon Progeny on Superconducting Qubits

Early Multiwavelength Observations of AT 2026fgk: The Luminous Afterglow to Sub-luminous GRB 260310A, Identified Independently of a Gamma-ray Trigger

Hasse-Weil Zeta Functions Modulo a Prime

Muon-induced di-tau production as a probe of new physics

Posterior-Aware Differential Channel Tracking for Reliable Single-Stream DAB+ Passive Radar

GRB 260310A/SN 2026fgk: Photometric and Spectroscopic Evolution of a Nearby GRB-Supernova and an Exceptionally Bright Afterglow at z=0.153

A Human-in-the-Loop Framework for Efficient Prompt Selection in Microscopy Vision-Language Models

Denoising Interferometric Observations Using Visibilities-Informed Neural Networks

Causality and Scientific Inquiry: Lessons from Space Physics and Medical Sciences

Towards Closed-loop Stability of Nonlinear Receding Horizon Games