AstroSpecLM: A Spectrum-Language Model for Evidence-Grounded Astronomical Spectral Analysis

Jinghang Shi; Yanxia Zhang; Ali Luo; Changhua Li; Xiao Kong

astro-ph.IMSep 7, 2026

AstroSpecLM: A Spectrum-Language Model for Evidence-Grounded Astronomical Spectral Analysis

Authors: Jinghang Shi, Yanxia Zhang, Ali Luo, Changhua Li, Xiao Kong

Abstract

Astronomical spectra encode rich physical information, but drawing scientific conclusions from spectral features typically requires expert interpretation. This paper presents AstroSpecLM, a spectrum-language model that connects one-dimensional DESI spectra with Qwen3-4B to answer questions and provide explanations grounded in spectral evidence. Instead of generating question-answer pairs directly from templates or raw catalog fields, we first distill each spectrum into a compact set of catalog- and spectrum-derived facts, then use these facts as references to generate instruction-following conversations. The resulting model is competitive with specialist supervised baselines on classification and redshift estimation, while additionally producing natural-language explanations that reference specific spectral features. Our results indicate that grounding a language model in one-dimensional scientific spectra is feasible, and that fact-mediated instruction data yields a model capable of both prediction and explanation.

Figures & tables

Figure 1: Instruction data construction pipeline. Raw DESI spectra are first distilled into grounded facts combining catalog annotations and spectral measurements; these facts then serve as references for generating instruction-following conversations.

Figure 2: Overview of AstroSpecLM. A normalized DESI spectrum is encoded by the SpecMAE encoder, compressed by a spectral projector, and inserted into Qwen3-4B as continuous spectral embeddings. During spectral instruction tuning, the SpecMAE encoder is frozen, while the projector, special-token embeddings, and LoRA adapters are trained.

Model	Spectrum-grounded Tasks				Language Benchmarks
	Object Class	Stellar Subtype	Redshift	$\boldsymbol{T_{\mathrm{eff}}}$	MMLU-Redux	IFEval	MATH-500
	Acc.	Acc.	MAE / R 2	MAE / R 2	Acc.	Acc.	Acc.
SpecMAE	98.12	85.62	0.029/ 0.993	162.3 / 0.955	N/A	N/A	N/A
AstroCLIP ( Parker et al., 2024 )	97.82	86.37	0.099/0.969	330.3/0.942	N/A	N/A	N/A
Qwen3-4B ( Yang et al., 2025 )	N/A	N/A	N/A	N/A	75.82	69.30	83.60
AstroSpecLM	98.56	87.56	0.015 /0.992	335.2/0.668	74.00	65.53	80.20

Table 1: Main results on spectrum-grounded tasks and language benchmarks. Best results are shown in bold. N/A indicates tasks that are not applicable due to model input or output constraints.

Figure 3: Qualitative examples comparing AstroSpecLM with SpecMAE and AstroCLIP. Left: redshift estimation for a QSO at $z=4.3602$ . Center: stellar subtype classification for a white dwarf. Right: a galaxy misclassified as a star by AstroSpecLM.

Model Variant	Object Class	Stellar Subtype	Redshift	$T_{\rm eff}$	MMLU-Redux
	Acc.	Acc.	MAE / R 2	MAE / R 2	Acc.
Training strategy
AstroSpecLM (4B, SFT)	98.56	87.56	0.015 / 0.992	335.2 / 0.668	74.00
+ Alignment pretraining	98.46	86.82	0.011 / 0.992	206.2 / 0.942	70.40
+ 2 training epochs	98.78	85.62	0.031 / 0.975	350.6 / 0.712	72.10
Model scale & token length

Table 2: Ablation results. The first row (4B, SFT, 130 tok, rank=16) is our final configuration. Best results are in bold, second-best results are underlined.

Appendix figures & tables7 assets

Supplementary material from the paper’s appendix.

Appendix

Fact Type	Fields	Source
Object labels	spectral type, subtype, redshift	catalog
Continuum	slope label, red-to-blue ratio	measured
D n 4000	value, strength label	measured
Spectral lines	rest/observed wavelength, EW, significance, status	measured
Molecular bands	CaH, TiO, VO indices (K/M stars)	measured
Stellar parameters	effective temperature (when available)	catalog

Appendix

Table 3: Structured fact types used as grounding references.

Object Class	Representative Spectral Features
GALAXY	[O II], [O III], H $\alpha$ , H $\beta$ , [N II], [S II], Ca H/K, Mg b, Na D
QSO	Ly $\alpha$ , N V, Si IV+O IV], C IV, C III], Mg II, H $\alpha$ , H $\beta$ , [O III]
STAR	Ca H/K, Balmer lines, Mg b, Na D, H $\alpha$ , Ca II triplet
K/M STAR	CaH, TiO molecular bands

Appendix

Table 4: Class-specific spectral lines and bands used in fact extraction.

Panel of Figure 4 — Figure 4: Full prompt used for instruction data generation.

Table 9

Panel of Figure 5 — Figure 5: Full prompt used to generate spectrum–caption pairs for alignment pretraining.

Hyperparameter	SpecMAE	AstroSpecLM
Training data	$\mathcal{D}_{\mathrm{enc}}$	DESI-SpecInstruct
Epochs	100	1
Batch size	360	32
Optimizer	AdamW	AdamW
Learning rate	$1\times 10^{-4}$	$2\times 10^{-5}$
Weight decay	0.001	0.01

Appendix

Table 7: Training hyperparameters for SpecMAE pretraining and AstroSpecLM spectral instruction tuning.

Panel of Figure 6 — Figure 6: Full prompt used for the LLM-as-a-judge faithfulness evaluation.

Explore similar work

Spectra as Language: Large Language Models for Scalable Stellar Parameter and Abundance Inference
May 21, 2026Hai-Ling Lu, Yu-Yang Li, Yin-Bi Li +4AI-Assisted Scientific Research
National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100101, China · University of Chinese Academy of Sciences, Beijing 100049, China · School of Astronomy and Space Science, University of Chinese Academy of Sciences, Beijing 100049, China +1
A systematic evaluation of vision-language models for observational astronomical reasoning tasks
Apr 27, 2026Wenke Ren, Hengxiao Guo, Wenwen Zuo +1VLM Evaluation Vision-Language Models
Shanghai Astronomical Observatory, Chinese Academy of Sciences, Shanghai, China · Harvard University, Boston, Massachusetts, USA
SpecVQA: A Benchmark for Spectral Understanding and Visual Question Answering in Scientific Images
Apr 30, 2026Jialu Shen, Han Lyu, Suyang Zhong +5Visual Question Answering Multimodal Model Evaluation
DP Technology

May 21, 2026J/K move · Enter open · S save

Spectra as Language: Large Language Models for Scalable Stellar Parameter and Abundance Inference

Hai-Ling Lu, Yu-Yang Li, Yin-Bi Li, Cun-Shi Wang+3

National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100101, China · University of Chinese Academy of Sciences, Beijing 100049, China · School of Astronomy and Space Science, University of Chinese Academy of Sciences, Beijing 100049, China+1

AI-Assisted Scientific Research

Open paper AskPDF arXiv

Stellar spectra encode key information on the physical properties and chemical compositions of stars. Accurate stellar parameter determination is essential for addressing major questions such as galaxy and stellar evolution. Large-scale spectroscopic surveys have accumulated unprecedented spectral data. Traditional feature extraction or model-fitting approaches struggle with high-dimensional, massive datasets, limited generalization, and computational inefficiency. Recent advances in large language models demonstrate strong generalization and feature-learning in tasks like natural language processing, DNA/RNA sequence analysis, and protein/chemical parsing. Stellar spectra are continuous sequential signals, enabling the transfer of language models to stellar spectroscopy. Here, we propose a two-stage large language model framework for stellar parameter inference, achieving accurate estimation of effective temperature, surface gravity, metallicity, and abundances of ~20 chemical elements. Scaling-law analyses show systematic performance improvements with increasing data, providing a scalable framework for forthcoming large-scale surveys.

Abstract

Explore similar work

Spectra as Language: Large Language Models for Scalable Stellar Parameter and Abundance Inference

A systematic evaluation of vision-language models for observational astronomical reasoning tasks

SpecVQA: A Benchmark for Spectral Understanding and Visual Question Answering in Scientific Images