Generalization from Low- to Moderate-Resolution Spectra with Neural Networks for Stellar Parameter Estimation: A Case Study with DESI

2026-02-16

AI Summary

This study addresses the challenge of estimating stellar parameters across heterogeneous spectroscopic surveys—specifically, transferring knowledge from low-resolution spectra (e.g., LAMOST) to medium-resolution data (e.g., DESI). The authors propose a transfer learning framework based on a multilayer perceptron (MLP), pre-trained on LAMOST spectra and systematically evaluated for zero-shot transfer and various fine-tuning strategies—including full-parameter fine-tuning, LoRA, and residual head adapters—on DESI data. Remarkably, the pre-trained MLP achieves strong performance without any fine-tuning, with modest improvements attainable through targeted adaptation. Comparative experiments reveal that MLPs operating directly on raw spectra excel for metal-poor stars, whereas Transformer-based embeddings yield superior iron abundance estimates for metal-rich stars, highlighting the surprising efficacy of simple MLP architectures in cross-survey spectral generalization.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Multiagent Systems: Multiagent Learning

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Cross-survey generalization is a critical challenge in stellar spectral analysis, particularly in cases such as transferring from low- to moderate-resolution surveys. We investigate this problem using pre-trained models, focusing on simple neural networks such as multilayer perceptrons (MLPs), with a case study transferring from LAMOST low-resolution spectra (LRS) to DESI medium-resolution spectra (MRS). Specifically, we pre-train MLPs on either LRS or their embeddings and fine-tune them for application to DESI stellar spectra. We compare MLPs trained directly on spectra with those trained on embeddings derived from transformer-based models (self-supervised foundation models pre-trained for multiple downstream tasks). We also evaluate different fine-tuning strategies, including residual-head adapters, LoRA, and full fine-tuning. We find that MLPs pre-trained on LAMOST LRS achieve strong performance, even without fine-tuning, and that modest fine-tuning with DESI spectra further improves the results. For iron abundance, embeddings from a transformer-based model yield advantages in the metal-rich ([Fe/H]>-1.0) regime, but underperform in the metal-poor regime compared to MLPs trained directly on LRS. We also show that the optimal fine-tuning strategy depends on the specific stellar parameter under consideration. These results highlight that simple pre-trained MLPs can provide competitive cross-survey generalization, while the role of spectral foundation models for cross-survey stellar parameter estimation requires further exploration.

Problem

• cross-survey generalization
• stellar parameter estimation
• low-resolution spectra
• moderate-resolution spectra
• neural networks

Innovation

• cross-survey generalization
• stellar parameter estimation
• pre-trained MLP
• spectral embeddings
• fine-tuning strategies