ESNet

Version: 2025.03.14 Added: 2025-06-20 Published: 2025-06-20 5.43M parameters Missing preds: 0
Repo Paper DOI Files PR Checkpoint
Predictions
Discovery

ML vs DFT Formation Energies

ML vs DFT Convex Hull Distance

Convex hull distance prediction errors projected onto elements

1 H 0.41
2 He 0.00
3 Li 0.18
4 Be 0.21
5 B 0.30
6 C 0.26
7 N 0.43
8 O 0.70
9 F 0.68
10 Ne 0.00
11 Na 0.20
12 Mg 0.18
13 Al 0.25
14 Si 0.35
15 P 0.33
16 S 0.54
17 Cl 0.56
18 Ar 0.00
19 K 0.19
20 Ca 0.19
21 Sc 0.21
22 Ti 0.36
23 V 0.51
24 Cr 0.28
25 Mn 0.27
26 Fe 0.33
27 Co 0.31
28 Ni 0.29
29 Cu 0.19
30 Zn 0.20
31 Ga 0.22
32 Ge 0.26
33 As 0.28
34 Se 0.41
35 Br 0.42
36 Kr 0.00
37 Rb 0.19
38 Sr 0.18
39 Y 0.18
40 Zr 0.27
41 Nb 0.38
42 Mo 0.28
43 Tc 0.17
44 Ru 0.22
45 Rh 0.22
46 Pd 0.21
47 Ag 0.18
48 Cd 0.17
49 In 0.21
50 Sn 0.23
51 Sb 0.20
52 Te 0.31
53 I 0.34
54 Xe 0.10
55 Cs 0.18
56 Ba 0.20
57 La 0.18
58 Ce 0.17
59 Pr 0.17
60 Nd 0.17
61 Pm 0.12
62 Sm 0.16
63 Eu 0.15
64 Gd 0.21
65 Tb 0.17
66 Dy 0.17
67 Ho 0.17
68 Er 0.17
69 Tm 0.18
70 Yb 0.16
71 Lu 0.18
72 Hf 0.25
73 Ta 0.48
74 W 0.19
75 Re 0.19
76 Os 0.20
77 Ir 0.24
78 Pt 0.23
79 Au 0.22
80 Hg 0.15
81 Tl 0.17
82 Pb 0.19
83 Bi 0.17
84 Po 0.00
85 At 0.00
86 Rn 0.00
87 Fr 0.00
88 Ra 0.00
89 Ac 0.10
90 Th 0.20
91 Pa 0.19
92 U 0.22
93 Np 0.23
94 Pu 0.34
95 Am 0.00
96 Cm 0.00
97 Bk 0.00
98 Cf 0.00
99 Es 0.00
100 Fm 0.00
101 Md 0.00
102 No 0.00
103 Lr 0.00
104 Rf 0.00
105 Db 0.00
106 Sg 0.00
107 Bh 0.00
108 Hs 0.00
109 Mt 0.00
110 Ds 0.00
111 Rg 0.00
112 Cn 0.00
113 Nh 0.00
114 Fl 0.00
115 Mc 0.00
116 Lv 0.00
117 Ts 0.00
118 Og 0.00
57-71 La-Lu Lanthanides
89-103 Ac-Lr Actinides

Model Authors

  1. Chao Huang  Institute of Computing Technology, Chinese Academy of Science, Beijing  
  2. Chunyan Chen  Ningbo Institute of Artificial Intelligence Industry, Ningbo, China  
  3. Ling Shi  Ningbo Institute of Artificial Intelligence Industry, Ningbo, China  

Trained By

  1. Ling Shi  Ningbo Institute of Artificial Intelligence Industry, Ningbo, China  

Model Info

  • Model Version 2025.03.14
  • Model Type Transformer
  • Targets E
  • Openness OSOD
  • Train Task RS2RE
  • Test Task IS2E
  • Trained for Benchmark Yes

Training Set

MP v2022.10.28: 155k structures

Description

ESNet is a graph neural network model designed for predicting the energy. The model builds on existing models based on crystal structure graph, to provide an in-depth analysis of how material elemental composition and crystal structure work together to influence material properties.

Hyperparameters

  • graph_construction_radius: 8
  • max_neighbors: 25

Dependencies