Mathematical & physical sciences · OECD FoS 1.3 · China MoE 0704 天文学
Astronomy & astrophysics
天文学
L2Local breakthroughs → Steady Score 2.45 · Last reviewed 2026-10-22 · Next review 2027-01-19
BIGGEST BREAKTHROUGH 2018-01-30 AI · majorField · major
Neural network finds Kepler-90i in Kepler data (AJ) ↗
Neural net finds Kepler-90i, tying the Solar System’s planet count
Why this level
ML finds exoplanet candidates and supernovae at scale (≈10,000 new TESS candidates, ApJS 2026; NASA ExoMiner++ ≈7,000), though most still need vetting and follow-up.
S = 0.30·method + 0.35·discovery + 0.25·adoption + 0.10·education (education scores 0 when there is no evidence; if empty, the other weights are renormalised)
Sources
- iopscience.iop.org/article/10.3847/1538-4365/ae5b6c
- science.nasa.gov/open-science/deep-learning-exoplanets-tess
Composite 1.5–2.49: at least one peer-reviewed/officially verified AI-key result, or one dimension at 3; routine workflow largely unchanged. All levels →
Research directions
Exoplanet searchReal-time transient classificationSurvey anomaly detection
Papers
- New substellar candidates identified through deep learning in the F150 sample of the large-scale SHINE direct imaging survey
AI: recovers all known companions and finds faint signals missed by classical pipelines. Field: candidates only; one (Smethells 20) remains strong
2026-10-15 · Astronomy & Astrophysics
- The T16 Planet Hunt: 10,000 New Planet Candidates from TESS Cycle 1 and the Confirmation of a Hot Jupiter around TIC 183374187*
~10,000 new candidates
2026-04-28 · The Astrophysical Journal Supplement Series
- The Dawes Review 10: The impact of deep learning for the analysis of galaxy surveys
Maps deep learning becoming a standard survey tool
2023-01-10 · Publications of the Astronomical Society of Australia
- ExoMiner: A Highly Accurate and Explainable Deep Learning Classifier That Validates 301 New Exoplanets
NASA neural network statistically validates 301 planets at once
2022-02-17 · The Astrophysical Journal
- Identifying Exoplanets with Deep Learning: A Five-planet Resonant Chain around Kepler-80 and an Eighth Planet around Kepler-90
Neural net finds Kepler-90i, tying the Solar System’s planet count
2018-01-30 · The Astronomical Journal
- Probabilistic Machine Learning for Short-Term Atmospheric Seeing Forecasting
Benchmarks probabilistic ML models for forecasting astronomical seeing up to two hours ahead, focusing on uncertainty quantification.
2026-10-21 · arXiv · Mary-Joe Medlej, Rahul Srinivasan, Simon Prunet, Aziz Ziad
- Data-driven characterisation of variable source light curves with multimodal representation learning
A survey-agnostic, label-free multimodal representation framework to characterise variable-source light curves from ZTF and Rubin LSST.
2026-10-21 · arXiv · M. Madarász, G. Marton, M. Audard
- A Hybrid Neural Network Emulator for M1 Neutrino Reaction Rates in Binary Neutron Star Merger Simulations
A hybrid neural emulator replaces only the costly 2D neutrino reactions in binary neutron star merger simulations.
2026-10-20 · arXiv · Yi Qiu, David Radice, Raffaele Ferrari, Albino Perego
- The "Rubin 2036: Opportunities for Discovery" Workshop: A Community Roadmap for the Future of the Vera C. Rubin Observatory After the LSST
Community roadmap for Rubin after LSST, including a foundation model pretrained on Rubin imaging for detection, deblending and classification.
2026-10-20 · arXiv · Andrés Alejandro Plazas Malagón, Steve Allen, Shreya Anand, Muhammad Awais
- J-PLUS: Identifying Outlier Spectral Energy Distributions in the Fourth Data Release Using Unsupervised Machine Learning
Autoencoder anomaly detection on 12-band J-PLUS photometry flags outlier SEDs for spectroscopic follow-up.
2026-10-19 · arXiv · F. Quispe-Huaynasi, F. Roig, V. Loaiza-Tacuri, V. M. Placco
- ASTROLABE: A 3D Particle-Level Surrogate for Cosmology Simulation with Physical Flow Matching
A flow-matching emulator acting on particle phase-space coordinates, replacing costly dark-matter N-body scans at high resolution.
2026-10-19 · arXiv · Ibrahim Elsharkawy, David Curtin, Yonatan Kahn, Neal Weiner
- Single-Transit Parameter Estimation with Calibrated Uncertainties
Automated parameter estimation with calibrated uncertainties for single-transit exoplanet events.
2026-10-17 · arXiv · Helem Salinas, Rafael Brahm, Greg Olmschenk, Melinda Soares-Furtado
- Attitude Estimation of Falcon 9 Rocket Body Based on Photometric Characteristics
Estimates the attitude of a Falcon 9 rocket body from photometric signatures for space-debris monitoring.
2026-10-17 · arXiv · Hui Li, Rong-Wang Li, Ru-Feng Tang, Yu-Qiang Li
- From arXiv, bioRxiv, medRxiv and OpenAlex (last ~30 days); summaries written by AIMap from the abstract.
Milestones 9
Big glowing nodes are leaps: AI leap cool · Field leap warm · double ring = both
-
Discovery
Deep learning (NA-SODINN) re-analysis of the SHINE F150 sample finds 13 new substellar candidates (A&A)AI: recovers all known companions and finds faint signals missed by classical pipelines. Field: candidates only; one (Smethells 20) remains strong
Astronomy & Astrophysics -
Paper
Amateur astronomer uses AI to find a possible hidden exoplanetAI tools let amateurs join frontier searches
Scientific American -
Dataset
Anthropic uses Claude to produce the first complete UV sky map, filling ~1/3 unobserved sky with calibrated predictionsAgent-built data product; predicted regions are not observations
AnthropicVendor claim -
Deployment
NOMAI: real-time superluminous-supernova classifier in Fink (A&A)ML in real-time alert brokering
A&ATo be verified -
Paper
Dawes Review: the impact of deep learning on galaxy-survey analysis (PASA)Maps deep learning becoming a standard survey tool
PASA -
Paper
ExoMiner validates 301 new Kepler exoplanets (ApJ)NASA neural network statistically validates 301 planets at once
The Astrophysical Journal -
Paper
Neural network finds Kepler-90i in Kepler data (AJ)Neural net finds Kepler-90i, tying the Solar System’s planet count
The Astronomical Journal
No milestones match this filter.
More papers & repos 5
Latest news 3
- Amateur astronomer harnesses AI to find a possible hidden exoplanetScientific Americanen