Inertial Confinement Fusion Forecasting via Large Language Models

TL;DR

A novel integration of Large Language Models with classical reservoir computing paradigms tailored to address a critical challenge, Laser-Plasma Instabilities, in Inertial Confinement Fusion, strives to forge an innovative synergy between AI and Inertial Confinement Fusion for advancing fusion energy.

Abstract

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

Figures (10)

• Figure 1: The overview of LPI-LLM.
• Figure 2: The ICF pipeline.
• Figure 3: LPI-LLMframework. (a) Fusion-specific prompts structure domain textual prompts with context, task and input descriptions (see §\ref{['sec:LLM-anchored Reservoir']} for details). (b) Signal-digesting channels comprise a pre-trained temporal encoder to extract time-series features of laser signals, and a spatial encoder to encode critical landscapes of the inputs (see §\ref{['sec:SDC']} for details). For simplicity, we skip some architectural modules. We provide implementation details in appendix (see §\ref{['Implementation']}).
• Figure 4: Pipeline of Confidence Scanner. We re-calibrate token confidence to align with the energy-prediction head.
• Figure 5: Qualitative results of 2 hot electron prediction cases. We plot Ground Truth and the predictions of Ours, Autoformer, Time-LLM and LSTM. Y and X axes denote hot electron energy in voltage, and time steps with step length of 0.025 nanosecond respectively.