Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Synthetic Graphs

Benchmarks

Task NameDataset NameSOTA ResultTrend
Causal DiscoverySynthetic Graphs (ER4, SF4) with Gaussian, Gumbel, and exponential noise
SHD4.28
15
Causal DiscoverySynthetic graphs N=50, E=50
F142.4
13
Runtime EfficiencySynthetic Graphs d=100
Runtime (s)0
12
Runtime EfficiencySynthetic Graphs d=50
Runtime (seconds)0
12
Runtime EfficiencySynthetic Graphs d=20
Runtime (seconds)0
12
Runtime EfficiencySynthetic Graphs d=10
Runtime (seconds)0
12
Runtime EfficiencySynthetic Graphs d=5
Runtime (s)0
12
Causal DiscoverySynthetic graphs N=100, E=100
SHD126.7
10
Causal DiscoverySynthetic graphs (N=20, E=30)
SHD26.7
7
Causal DiscoverySynthetic graphs (N=20, E=20)
Structural Hamming Distance (SHD)16.3
7
Subgraph CountingSynthetic graphs (test)
MAE (Triangle)0.0118
7
Skeleton RecoverySynthetic Graphs (d=500, n=1000)
Time (s)170.4
3
Skeleton RecoverySynthetic Graphs d=200, n=1000
Time (s)25.3
3
Skeleton RecoverySynthetic Graphs (d=100, n=1000)
Time (s)14.1
3
Discrete Stochastic Dynamics LearningSynthetic Graphs (N=6, beta=0.20)
Hellinger Distance0.069
2
Discrete Stochastic Dynamics LearningSynthetic Graphs (N=6, beta=0.10)
Hellinger Distance0.059
2
Discrete Stochastic Dynamics LearningSynthetic Graphs (N=6, beta=0.01)
Hellinger Distance0.066
2
Showing 17 of 17 rows