Check leaderboards for
The bold method name indicates that the implementation is official (by the author of the original paper).
Package denotes the required package version for each dataset to be eligible for the leaderboard.
Please update your package >= 0.7.5. In version 2 of tgbl-wiki, we have included all possible negative destinations for each positive edge thus increasing the task difficulty. For submissions, please submit to this version. Previous version of the dataset leaderboard is tgbl-wiki-v1. The MRR score on the test and validation sets. The higher, the better.
Package: >=0.7.5
| Rank |
Method |
Test MRR |
Validation MRR |
Contact |
References |
Date |
| 1 |
DyGFormer |
0.798 ± 0.004 |
0.816 ± 0.005 |
Le Yu (Beihang University) |
Paper, Code |
August 22nd, 2023 |
| 2 |
NAT |
0.749 ± 0.010 |
0.773 ± 0.011 |
TGB team |
Paper, Code |
August 22nd, 2023 |
| 3 |
TGR-TNCN |
0.724 ± 0.011 |
0.751 ± 0.005 |
Katarina Petrović (University of Oxford) |
Paper, Code |
October 23rd, 2024 |
| 4 |
TNCN |
0.718 ± 0.001 |
0.741 ± 0.001 |
Xiaohui Zhang (PKU) |
Paper, Code |
August 25th, 2024 |
| 5 |
CAWN |
0.711 ± 0.006 |
0.743 ± 0.004 |
TGB team |
Paper, Code |
July 27th, 2023 |
| 6 |
TGR-TGN |
0.589 ± 0.048 |
0.642 ± 0.047 |
Katarina Petrović (University of Oxford) |
Paper, Code |
October 23rd, 2024 |
| 7 |
EdgeBank(tw) |
0.571 |
0.600 |
TGB team |
Paper, Code |
July 27th, 2023 |
| 8 |
EdgeBank(unlimited) |
0.495 |
0.527 |
TGB team |
Paper, Code |
July 27th, 2023 |
| 9 |
TGN |
0.396 ± 0.060 |
0.435 ± 0.069 |
TGB team |
Paper, Code |
August 7th, 2023 |
| 10 |
TCL |
0.207 ± 0.025 |
0.198 ± 0.016 |
TGB team |
Paper, Code |
July 27th, 2023 |
| 11 |
TGAT |
0.141 ± 0.007 |
0.131 ± 0.008 |
TGB team |
Paper, Code |
August 22th, 2023 |
| 12 |
GraphMixer |
0.118 ± 0.002 |
0.113 ± 0.003 |
TGB team |
Paper, Code |
August 7th, 2023 |
| 13 |
DyRep |
0.050 ± 0.017 |
0.072 ± 0.009 |
TGB team |
Paper, Code |
July 27th, 2023 |
Please update your package >= 0.7.5. In version 2 of tgbl-review, we have included 100 negative destinations for each positive edge thus increasing the task difficulty. For submissions, please submit to this version. Previous version of the dataset leaderboard is tgbl-review-v1. The MRR score on the test and validation sets. The higher, the better.
Package: >=0.7.5
| Rank |
Method |
Test MRR |
Validation MRR |
Contact |
References |
Date |
| 1 |
TGR-TGN |
0.854 ± 0.011 |
0.834 ± 0.012 |
Katarina Petrović (University of Oxford) |
Paper, Code |
October 23rd, 2024 |
| 2 |
TGR-TNCN |
0.599 ± 0.045 |
0.511 ± 0.054 |
Katarina Petrović (University of Oxford) |
Paper, Code |
October 23rd, 2024 |
| 3 |
GraphMixer |
0.521 ± 0.015 |
0.428 ± 0.019 |
TGB team |
Paper, Code |
July 27th, 2023 |
| 4 |
TNCN |
0.377 ± 0.010 |
0.325 ± 0.003 |
Xiaohui Zhang (PKU) |
Paper, Code |
August 25th, 2024 |
| 5 |
TGAT |
0.355 ± 0.012 |
0.324 ± 0.006 |
TGB team |
Paper, Code |
August 22th, 2023 |
| 6 |
TGN |
0.349 ± 0.020 |
0.313 ± 0.012 |
TGB team |
Paper, Code |
July 27th, 2023 |
| 7 |
NAT |
0.341 ± 0.020 |
0.302 ± 0.011 |
TGB team |
Paper, Code |
August 22nd, 2023 |
| 8 |
DyGFormer |
0.224 ± 0.015 |
0.219 ± 0.017 |
Le Yu (Beihang University) |
Paper, Code |
September 21st, 2023 |
| 9 |
DyRep |
0.220 ± 0.030 |
0.216 ± 0.031 |
TGB team |
Paper, Code |
July 27th, 2023 |
| 10 |
CAWN |
0.193 ± 0.001 |
0.200 ± 0.001 |
TGB team |
Paper, Code |
July 27th, 2023 |
| 11 |
TCL |
0.193 ± 0.009 |
0.199 ± 0.007 |
TGB team |
Paper, Code |
July 27th, 2023 |
| 12 |
EdgeBank(tw) |
0.025 |
0.024 |
TGB team |
Paper, Code |
July 27th, 2023 |
| 13 |
EdgeBank(unlimited) |
0.023 |
0.023 |
TGB team |
Paper, Code |
July 27th, 2023 |
Please update your package >= 0.9.0. In version 2 of tgbl-coin, we fixed an error where a small portion of edges are not sorted chronologically. For submissions, please submit to this version. Previous version of the dataset leaderboard is tgbl-coin-v1. The MRR score on the test and validation sets. The higher, the better.
Package: >=0.9.0
| Rank |
Method |
Test MRR |
Validation MRR |
Contact |
References |
Date |
| 1 |
TGR-TNCN |
0.783 ± 0.033 |
0.769 ± 0.025 |
Katarina Petrović (University of Oxford) |
Paper, Code |
October 23rd, 2024 |
| 2 |
TNCN |
0.762 ± 0.004 |
0.740 ± 0.002 |
Xiaohui Zhang (PKU) |
Paper, Code |
August 25th, 2024 |
| 3 |
TGR-TGN |
0.755 ± 0.088 |
0.734 ± 0.082 |
Katarina Petrović (University of Oxford) |
Paper, Code |
October 23rd, 2024 |
| 4 |
DyGFormer |
0.752 ± 0.004 |
0.730 ± 0.002 |
Le Yu (Beihang University) |
Paper, Code |
November 2nd, 2023 |
| 5 |
TGN |
0.586 ± 0.037 |
0.607 ± 0.014 |
TGB team |
Paper, Code |
September 27th, 2023 |
| 6 |
EdgeBank(tw) |
0.580 |
0.492 |
TGB team |
Paper, Code |
September 27th, 2023 |
| 7 |
DyRep |
0.452 ± 0.046 |
0.512 ± 0.014 |
TGB team |
Paper, Code |
September 27th, 2023 |
| 8 |
EdgeBank(unlimited) |
0.359 |
0.315 |
TGB team |
Paper, Code |
September 27th, 2023 |
The MRR score on the test and validation sets. The higher, the better.
Package: >=0.7.5
| Rank |
Method |
Test MRR |
Validation MRR |
Contact |
References |
Date |
| 1 |
TGR-TNCN |
0.891 ± 0.056 |
0.899 ± 0.014 |
Katarina Petrović (University of Oxford) |
Paper, Code |
October 23rd, 2024 |
| 2 |
TGR-TGN |
0.836 ± 0.079 |
0.826 ± 0.091 |
Katarina Petrović (University of Oxford) |
Paper, Code |
October 23rd, 2024 |
| 3 |
TNCN |
0.697 ± 0.006 |
0.643 ± 0.003 |
Xiaohui Zhang (PKU) |
Paper, Code |
August 25th, 2024 |
| 4 |
DyGFormer |
0.670 ± 0.001 |
0.613 ± 0.003 |
Le Yu (Beihang University) |
Paper, Code |
September 21st, 2023 |
| 5 |
TGN |
0.379 ± 0.021 |
0.356 ± 0.019 |
TGB team |
Paper, Code |
June 7th, 2023 |
| 6 |
DyRep |
0.289 ± 0.033 |
0.291 ± 0.028 |
TGB team |
Paper, Code |
June 7th, 2023 |
| 7 |
EdgeBank(tw) |
0.149 |
0.124 |
TGB team |
Paper, Code |
June 7th, 2023 |
| 8 |
EdgeBank(unlimited) |
0.129 |
0.109 |
TGB team |
Paper, Code |
June 7th, 2023 |
Please update your package >= 0.9.2. In version 2 of tgbl-flight, we fixed an error where a small portion of edges are not sorted chronologically. For submissions, please submit to this version. Previous version of the dataset leaderboard is tgbl-flight-v1. The MRR score on the test and validation sets. The higher, the better.
Package: >=0.9.2
| Rank |
Method |
Test MRR |
Validation MRR |
Contact |
References |
Date |
| 1 |
TGR-TNCN |
0.851 ± 0.011 |
0.854 ± 0.014 |
Katarina Petrović (University of Oxford) |
Paper, Code |
October 23rd, 2024 |
| 2 |
TNCN |
0.820 ± 0.004 |
0.831 ± 0.003 |
Xiaohui Zhang (PKU) |
Paper, Code |
August 25th, 2024 |
| 3 |
TGR-TGN |
0.814 ± 0.023 |
0.818 ± 0.018 |
Katarina Petrović (University of Oxford) |
Paper, Code |
October 23rd, 2024 |
| 4 |
TGN |
0.705 ± 0.020 |
0.731 ± 0.010 |
TGB team |
Paper, Code |
September 27th, 2023 |
| 5 |
DyRep |
0.556 ± 0.014 |
0.573 ± 0.013 |
TGB team |
Paper, Code |
September 27th, 2023 |
| 6 |
EdgeBank(tw) |
0.387 |
0.363 |
TGB team |
Paper, Code |
September 27th, 2023 |
| 7 |
EdgeBank(unlimited) |
0.167 |
0.166 |
TGB team |
Paper, Code |
September 27th, 2023 |