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

GraphWalks

Benchmarks

Task NameDataset NameSOTA ResultTrend
Synthetic Long-context ReasoningGraphWalks
Accuracy68.91
12
Long Context PerformanceGraphwalks Bfs (<128K)
Score98
9
Graph traversalGraphWalks
Parents Score89.87
8
Long-context graph reasoningGraphWalks
F1 Score82.3
6
Long Context PerformanceGraphwalks Parents
Score99.7
5
Long Context PerformanceGraphwalks Parents (<128K)
Score100
4
Showing 6 of 6 rows