benchmark
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Small Benchmark
12.407Shape
13
Original benchmark B
61.26Score
13
Zero-shot cross-domain benchmark (test)
5.94Mean
12
benchmark (test)
0.561Style Similarity CSD Score
9
70-example benchmark 1.0 (test)
0.59FaceSim Arc
9
Benchmark 17x256x256 resolution (test)
210.9gFVD
9
100-task benchmark (test)
0PE (cm)
8
33-task benchmark Survival prediction
58C-index
8
Small-scale benchmark Overall
33VR
8
Benchmark of 52 prompts and 20 style images 1.0 (test)
0.235Text Alignment
8
Benchmark 03
84In-Scope Accuracy
8
10-task benchmark Overall
0.02Demo L (MSE)
7
Benchmark II Waypoint-Hold Under Directionally Varied Push
3WP Success Count
7
benchmark local images disk-replay protocol (n=120) v6
518TTFT
6
14-task benchmark
95.6Final SR
6
200-task benchmark
599Success Rate
6
Benchmark ¬ATOMIC
83Valid Precision
6
8-task benchmark
94.8ID Score
6
Benchmark 1 (test)
97M_C (Completeness/Score)
6
benchmark High-Quality (HQ) 1.0
1.58Median Error (mm)
6
benchmark_1000 n=993 (train)
79.6Accuracy
5
Benchmark 2
17.2Model Complexity
5
Benchmark (BM) 10 clients, pathological non-IID
91AUC-ROC
5
benchmark Small Instances
263.33Objective Value
5
Benchmark Second Turn
2.32Block Efficiency
5