Inertial Confinement Fusion Forecasting via Large Language Models

Controlled fusion energy is deemed pivotal for the advancement of human civilization. In this study, we introduce LPI-LLM, a novel integration of Large Language Models (LLMs) with classical reservoir computing paradigms tailored to address a critical challenge, Laser-Plasma Instabilities (textttLPI), in Inertial Confinement Fusion (textttICF). Our approach offers several key contributions: Firstly, we propose the textitLLM-anchored Reservoir, augmented with a textitFusion-specific Prompt, enabling accurate forecasting of textttLPI-generated-hot electron dynamics during implosion. Secondly, we develop textitSignal-Digesting Channels to temporally and spatially describe the driver laser intensity across time, capturing the unique characteristics of textttICF inputs. Lastly, we design the textitConfidence Scanner to quantify the confidence level in forecasting, providing valuable insights for domain experts to design the textttICF process. Extensive experiments demonstrate the superior performance of our method, achieving 1.90 CAE, 0.14 texttttop-1 MAE, and 0.11 texttttop-5 MAE in predicting Hard X-ray (textttHXR) energies emitted by the hot electrons in textttICF implosions, which presents state-of-the-art comparisons against concurrent best systems. Additionally, we present LPI4AI, the first textttLPI benchmark based on physical experiments, aimed at fostering novel ideas in textttLPI research and enhancing the utility of LLMs in scientific exploration. Overall, our work strives to forge an innovative synergy between AI and textttICF for advancing fusion energy.