Deep learning-enhanced design for functionally graded auxetic lattices

•An efficient DNN-based design framework for functionally graded lattices is established.•Functionally graded lattices are designed using the framework, fabricated using addictive manufacturing, and evaluated using experimental tests.•The effectiveness of the design framework is demonstrated with the performances of the functionally graded lattices.•This work is inspirable to promote advanced structures/materials with tailored properties.