Corrigendum to “Thermodynamics-informed multi-head attention neural networks for constitutive modelling” Int. J. Solids Struct. 328 (2026) 113845

Key Points

• Thermodynamics-informed models improve performance in constitutive modelling tasks.
• Key adjustments made in neural network algorithms enhance predictive capabilities.
• Observational analysis facilitates a deeper understanding of material behaviors under different conditions.
• Highlights the importance of refining computational models to ensure accuracy and reliability.