arXiv subjects

Andrew Dunn

Publications and source records attributed to Andrew Dunn.

2 recordsLinked to original sources

arXiv2026

HARMONI at ELT: Reoptimization of the NGSS natural guide star sensors system for MCAO compatibility

HARMONI is the first-light, adaptive optics assisted, near-IR integral field spectrograph for the ELT. It covers a spectral range from below 800nm to 2400nm with resolving powers from 3000 to 7000 and spatial sampling of 25mas and 6mas. It can operate in three adaptive optics (AO) modes: single conjugate AO (SCAO), high-contrast AO (HCAO), and multi-conjugate AO (MCAO). The project is resuming its final design phase after a rescope design phase that has lasted two years. This paper describes the impact of the rescope on the natural guide star sensors (NGSS) system, imposing wide-ranging changes in particular to the low-order wavefront sensors (LOWFS) previously optimized for LTAO operation as well as on the external support and enclosure (ESE) whose external interfaces have changed. The high-contrast module (HCM) is also deeply affected due to the modified plate scale of the instrument, and the single-conjugate AO sensor (SCAOS) benefits from a reduced patrol field requirement. Description of the changes will be given together with the associated system analysis involving requirements flow-down, revisited functional analysis, and updated product breakdown structure. Elements of the new system management structure and work breakdown will also be included.

astro-ph.IM↗

arXiv2021

Context-Sensitive Visualization of Deep Learning Natural Language Processing Models

The introduction of Transformer neural networks has changed the landscape of Natural Language Processing (NLP) during the last years. So far, none of the visualization systems has yet managed to examine all the facets of the Transformers. This gave us the motivation of the current work. We propose a new NLP Transformer context-sensitive visualization method that leverages existing NLP tools to find the most significant groups of tokens (words) that have the greatest effect on the output, thus preserving some context from the original text. First, we use a sentence-level dependency parser to highlight promising word groups. The dependency parser creates a tree of relationships between the words in the sentence. Next, we systematically remove adjacent and non-adjacent tuples of \emph{n} tokens from the input text, producing several new texts with those tokens missing. The resulting texts are then passed to a pre-trained BERT model. The classification output is compared with that of the full text, and the difference in the activation strength is recorded. The modified texts that produce the largest difference in the target classification output neuron are selected, and the combination of removed words are then considered to be the most influential on the model's output. Finally, the most influential word combinations are visualized in a heatmap.

cs.CL↗