Exoformer: Accelerating Bayesian atmospheric retrievals with transformer neural networks

L. Pagliaro; T. Zingales; G. Piotto; I. Giovannini; G. Mantovan

doi:10.1051/0004-6361/202558264/pdf

March 26, 2026

Exoformer: Accelerating Bayesian atmospheric retrievals with transformer neural networks

View at Publisher

Ask AI

Bookmark

Share

Key Points

The research aims to enhance the efficiency of Bayesian atmospheric retrievals for exoplanetary spectra using transformer neural networks.
Developed a transformer-based neural network for generating informative prior distributions.
Applied the method to both simulated and JWST observations of WASP-39b and WASP-17b.
Integrated the new priors into the TauREx retrieval framework utilizing the nested sampling algorithm.
Achieved acceleration of nested-sampling retrievals by a factor of 3-8 compared to standard methods.
Preserved the consistency of retrieved parameters and best-fit spectra with classical techniques.
Confirmed statistical consistency between the hybrid approach and traditional methods with log-Bayesian evidence showing no strong preference.

No abstract yet

We only have this paper's title and publication details.

Read it at the source

Implication

Demonstrates a novel approach to improve atmospheric retrieval efficiency in exoplanets by leveraging transformer neural networks, suggesting significant advancements in data analysis.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Pagliaro et al. (2026) studied this question.

synapsesocial.com/papers/69e3215140886becb654087b https://doi.org/10.1051/0004-6361/202558264/pdf

View at Publisher

Ask AI

Bookmark

Share

Exoformer: Accelerating Bayesian atmospheric retrievals with transformer neural networks

Key Points

No abstract yet

Implication

Discussion

Cite This Study

Exoformer: Accelerating Bayesian atmospheric retrievals with transformer neural networks

Key Points

No abstract yet

Implication

Discussion

Cite This Study

Keep reading

More related papers

Keep reading

More related papers