Pre-trained MLPs on LAMOST low-resolution spectra generalize to DESI medium-resolution spectra for [Fe/H] and [α/Fe], outperforming the DESI SP pipeline in zero-shot and improving with modest fine-tuning.
SPARCL: SPectra Analysis and Retrievable Catalog Lab
1 Pith paper cite this work, alongside 2 external citations. Polarity classification is still indexing.
1
Pith paper citing it
2
external citations · Pith
abstract
SPectra Analysis and Retrievable Catalog Lab (SPARCL) at NOIRLab's Astro Data Lab was created to efficiently serve large optical and infrared spectroscopic datasets. It consists of services, tools, example workflows and currently contains spectra for over 7.5 million stars, galaxies and quasars from the Sloan Digital Sky Survey (SDSS) and the Dark Energy Spectroscopic Instrument (DESI) survey. We aim to eventually support the broad range of spectroscopic datasets that will be hosted at NOIRLab and beyond. Major elements of SPARCL include capabilities to discover and query for spectra based on parameters of interest, a fast web service that delivers desired spectra either individually or in bulk as well as documentation and example Jupyter Notebooks to empower users in their research. More information is available on the SPARCL website (https://astrosparcl.datalab.noirlab.edu).
fields
astro-ph.SR 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Generalization from Low- to Moderate-Resolution Spectra with Neural Networks for Stellar Parameter Estimation: A Case Study with DESI
Pre-trained MLPs on LAMOST low-resolution spectra generalize to DESI medium-resolution spectra for [Fe/H] and [α/Fe], outperforming the DESI SP pipeline in zero-shot and improving with modest fine-tuning.