エピソード

  • AI weather models independently learned physics

    AI weather models independently learned physics
    2026/08/30
    Large data-driven physics models like DeepMind's weather model GraphCast have empirically succeeded in parameterizing time operators for complex dynamical systems with an accuracy reaching or in some cases exceeding that of traditional physics-based solvers. Unfortunately, how these data-driven models perform computations is largely unknown and whether their internal representations are interpretable or physically consistent is an open question. Here, the authors adapt tools from interpretability research in Large Language Models to analyze intermediate computational layers in GraphCast, leveraging sparse autoencoders to discover interpretable features in the neuron space of the model. They uncover distinct features on a wide range of length and time scales that correspond to tropical cyclones, atmospheric rivers, diurnal and seasonal behavior, large-scale precipitation patterns, specific geographical coding, and sea-ice extent, among others. They further demonstrate how the precise abstraction of these features can be probed via interventions on the prediction steps of the model. As a case study, they sparsely modify a feature corresponding to tropical cyclones in GraphCast and observe interpretable and physically consistent modifications to evolving hurricanes. Such methods offer a window into the black-box behavior of data-driven physics models and are a step towards realizing their potential as trustworthy predictors and scientifically valuable tools for discovery. Paper: https://arxiv.org/abs/2512.24440
    続きを読む 一部表示
    42 分
  • GraphCast learns the laws of physics

    GraphCast learns the laws of physics
    2026/08/29
    Large data-driven physics models like DeepMind's weather model GraphCast have empirically succeeded in parameterizing time operators for complex dynamical systems with an accuracy reaching or in some cases exceeding that of traditional physics-based solvers. Unfortunately, how these data-driven models perform computations is largely unknown and whether their internal representations are interpretable or physically consistent is an open question. Here, the authors adapt tools from interpretability research in Large Language Models to analyze intermediate computational layers in GraphCast, leveraging sparse autoencoders to discover interpretable features in the neuron space of the model. They uncover distinct features on a wide range of length and time scales that correspond to tropical cyclones, atmospheric rivers, diurnal and seasonal behavior, large-scale precipitation patterns, specific geographical coding, and sea-ice extent, among others. They further demonstrate how the precise abstraction of these features can be probed via interventions on the prediction steps of the model. As a case study, they sparsely modify a feature corresponding to tropical cyclones in GraphCast and observe interpretable and physically consistent modifications to evolving hurricanes. Such methods offer a window into the black-box behavior of data-driven physics models and are a step towards realizing their potential as trustworthy predictors and scientifically valuable tools for discovery. Paper: https://arxiv.org/abs/2512.24440
    続きを読む 一部表示
    20 分
  • Walrus masters physics from stars to oceans

    Walrus masters physics from stars to oceans
    2026/08/16
    Foundation models have reshaped language and vision, but physical simulation has resisted the same playbook: heterogeneous data, unstable long-term rollouts, and mismatched resolutions and dimensionalities make it hard to train one model across many kinds of physics. This paper introduces Walrus, a transformer-based foundation model for fluid-like continuum dynamics, built around a harmonic-analysis-based stabilization method, load-balanced distributed 2D and 3D training strategies, and compute-adaptive tokenization. Walrus is pretrained on nineteen diverse scenarios spanning astrophysics, geoscience, rheology, plasma physics, acoustics, and classical fluids, treating Earth's atmosphere and oceans as just one flavor of continuum dynamics among many. Across short- and long-horizon forecasts on downstream tasks, and across the full breadth of pretraining domains, Walrus outperforms prior foundation models, while ablation studies confirm that the stabilization, distributed training, and tokenization choices each meaningfully improve forecast stability, training throughput, and transfer performance. Code and weights are released for community use. Paper: https://arxiv.org/abs/2511.15684
    続きを読む 一部表示
    17 分
  • Walrus AI unifies nineteen physics scenarios

    Walrus AI unifies nineteen physics scenarios
    2026/08/09
    Foundation models have reshaped language and vision, but physical simulation has resisted the same playbook: heterogeneous data, unstable long-term rollouts, and mismatched resolutions and dimensionalities make it hard to train one model across many kinds of physics. This paper introduces Walrus, a transformer-based foundation model for fluid-like continuum dynamics, built around a harmonic-analysis-based stabilization method, load-balanced distributed 2D and 3D training strategies, and compute-adaptive tokenization. Walrus is pretrained on nineteen diverse scenarios spanning astrophysics, geoscience, rheology, plasma physics, acoustics, and classical fluids, treating Earth's atmosphere and oceans as just one flavor of continuum dynamics among many. Across short- and long-horizon forecasts on downstream tasks, and across the full breadth of pretraining domains, Walrus outperforms prior foundation models, while ablation studies confirm that the stabilization, distributed training, and tokenization choices each meaningfully improve forecast stability, training throughput, and transfer performance. Code and weights are released for community use. Paper: https://arxiv.org/abs/2511.15684
    続きを読む 一部表示
    22 分
  • Teaching AI the laws of physics

    Teaching AI the laws of physics
    2026/08/02
    Foundation models have reshaped language and vision, but physical simulation has resisted the same playbook: heterogeneous data, unstable long-term rollouts, and mismatched resolutions and dimensionalities make it hard to train one model across many kinds of physics. This paper introduces Walrus, a transformer-based foundation model for fluid-like continuum dynamics, built around a harmonic-analysis-based stabilization method, load-balanced distributed 2D and 3D training strategies, and compute-adaptive tokenization. Walrus is pretrained on nineteen diverse scenarios spanning astrophysics, geoscience, rheology, plasma physics, acoustics, and classical fluids, treating Earth's atmosphere and oceans as just one flavor of continuum dynamics among many. Across short- and long-horizon forecasts on downstream tasks, and across the full breadth of pretraining domains, Walrus outperforms prior foundation models, while ablation studies confirm that the stabilization, distributed training, and tokenization choices each meaningfully improve forecast stability, training throughput, and transfer performance. Code and weights are released for community use. Paper: https://arxiv.org/abs/2511.15684
    続きを読む 一部表示
    19 分
  • Aurora AI beats traditional weather models

    Aurora AI beats traditional weather models
    2026/07/26
    Aurora is a pioneering AI foundation model created by Microsoft to revolutionize Earth system forecasting by delivering high-resolution predictions with unprecedented speed and efficiency. Developed using a 3D Swin Transformer architecture, the model was pretrained on over one million hours of diverse geophysical data to learn the complex dynamics of the atmosphere. Through a process of fine-tuning, Aurora can be specialized for diverse tasks, including weather forecasting, air pollution tracking, ocean wave modeling, and tropical cyclone tracking. Research indicates that it consistently matches or outperforms state-of-the-art numerical models, such as those from the ECMWF, while operating at a computational cost orders of magnitude lower than traditional simulations. The most recent iteration, Aurora 1.5, introduces expanded variables and probabilistic ensemble forecasting to better quantify atmospheric uncertainty. By providing an open-source framework via GitHub, Microsoft aims to democratize access to accurate environmental data and support global efforts in disaster mitigation and climate science. Link: https://www.microsoft.com/en-us/research/project/aurora-forecasting/
    続きを読む 一部表示
    19 分
  • Microsoft Aurora outperforms traditional weather models

    Microsoft Aurora outperforms traditional weather models
    2026/07/26
    Aurora is a pioneering AI foundation model created by Microsoft to revolutionize Earth system forecasting by delivering high-resolution predictions with unprecedented speed and efficiency. Developed using a 3D Swin Transformer architecture, the model was pretrained on over one million hours of diverse geophysical data to learn the complex dynamics of the atmosphere. Through a process of fine-tuning, Aurora can be specialized for diverse tasks, including weather forecasting, air pollution tracking, ocean wave modeling, and tropical cyclone tracking. Research indicates that it consistently matches or outperforms state-of-the-art numerical models, such as those from the ECMWF, while operating at a computational cost orders of magnitude lower than traditional simulations. The most recent iteration, Aurora 1.5, introduces expanded variables and probabilistic ensemble forecasting to better quantify atmospheric uncertainty. By providing an open-source framework via GitHub, Microsoft aims to democratize access to accurate environmental data and support global efforts in disaster mitigation and climate science. Link: https://www.microsoft.com/en-us/research/project/aurora-forecasting/
    続きを読む 一部表示
    24 分
  • Aurora AI beats multi-billion dollar supercomputers

    Aurora AI beats multi-billion dollar supercomputers
    2026/07/26
    Aurora is a pioneering AI foundation model created by Microsoft to revolutionize Earth system forecasting by delivering high-resolution predictions with unprecedented speed and efficiency. Developed using a 3D Swin Transformer architecture, the model was pretrained on over one million hours of diverse geophysical data to learn the complex dynamics of the atmosphere. Through a process of fine-tuning, Aurora can be specialized for diverse tasks, including weather forecasting, air pollution tracking, ocean wave modeling, and tropical cyclone tracking. Research indicates that it consistently matches or outperforms state-of-the-art numerical models, such as those from the ECMWF, while operating at a computational cost orders of magnitude lower than traditional simulations. The most recent iteration, Aurora 1.5, introduces expanded variables and probabilistic ensemble forecasting to better quantify atmospheric uncertainty. By providing an open-source framework via GitHub, Microsoft aims to democratize access to accurate environmental data and support global efforts in disaster mitigation and climate science. Link: https://www.microsoft.com/en-us/research/project/aurora-forecasting/
    続きを読む 一部表示
    45 分