TokaMark: A Comprehensive Benchmark for MAST Tokamak Plasma Models

TL;DR

TokaMark addresses the lack of open, standardized benchmarks for AI models in fusion plasma by introducing a comprehensive, open benchmark built on FAIR-MAST data from the MAST tokamak. It defines 14 downstream tasks across four groups to probe representation learning, temporal reasoning across fast and slow dynamics, robustness to incomplete data, and generalization across operating regimes, all within a hierarchical evaluation framework. The paper provides a baseline multi-branch encoder–decoder model, a detailed data preparation pipeline, and open-source tooling to enable fair cross-model comparisons and reproducible progress. By unifying multi-modal diagnostics and offering structured evaluation, TokaMark aims to accelerate data-driven plasma AI development and support safer, more reliable fusion operation and eventual commercialization.

Abstract

Development and operation of commercially viable fusion energy reactors such as tokamaks require accurate predictions of plasma dynamics from sparse, noisy, and incomplete sensors readings. The complexity of the underlying physics and the heterogeneity of experimental data pose formidable challenges for conventional numerical methods, while simultaneously highlights the promise of modern data-native AI approaches. A major obstacle in realizing this potential is, however, the lack of curated, openly available datasets and standardized benchmarks. Existing fusion datasets are scarce, fragmented across institutions, facility-specific, and inconsistently annotated, which limits reproducibility and prevents a fair and scalable comparison of AI approaches. In this paper, we introduce TokaMark, a structured benchmark to evaluate AI models on real experimental data collected from the Mega Ampere Spherical Tokamak (MAST). TokaMark provides a comprehensive suite of tools designed to (i) unify access to multi-modal heterogeneous fusion data (ii) harmonize formats, metadata, temporal alignment and evaluation protocols to enable consistent cross-model and cross-task comparisons. The benchmark includes a curated list of 14 tasks spanning a range of physical mechanisms, exploiting a variety of diagnostics and covering multiple target use cases. A baseline model is provided to facilitate transparent comparison and validation within a unified framework. By establishing a unified benchmark for both the fusion and AI-for-science communities, TokaMark aims to accelerate progress in data-driven plasma AI modeling, contributing to the broader goal of achieving sustainable and stable fusion energy. The benchmark, documentation, and tooling will be fully open sourced upon acceptance to encourage community adoption and contribution.

Figures (2)

• Figure 1: Examples of multi-modal signals from the FAIR-MAST dataset: a) Time series of plasma current, beta normal, NBI power, mirnov coils, and D$_\text{alpha}$ signals; b) Thomson scattering profiles of electron temperature, density, and pressure; c) maps of plasma current and magnetic flux; d) profile of the mirnov coils spectrogram.
• Figure 2: Multi-branch convolutional encoder–decoder