benchmark suite
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| General Language Understanding | 10-benchmark suite average | Average Accuracy73.81 | 15 | |
| Generalization across multiple tasks | Full Benchmark Suite Aggregate | Average Accuracy82.2 | 10 | |
| Offline Policy Adaptation | Full Benchmark Suite Gravity and Morph shifts | Total Score820.8 | 7 | |
| Visual instruction tuning | Benchmark Suite Aggregate | Average Score65.1 | 6 | |
| Quantum architecture compilation | 36-circuit benchmark suite Combined | Geometric-mean Execution Time19,507.5 | 4 | |
| Semantic Segmentation | 20-domain benchmark suite | h-mean58.05 | 4 | |
| Text Generation | benchmark suite 20-task | Average End-to-End Latency (ms)2,520 | 3 | |
| Unique bug detection | Full benchmark suite Very Large | TP6 | 3 | |
| LLM Pruning Evaluation | Full Benchmark Suite | S_tot Retention75.15 | 2 |
Showing 9 of 9 rows