| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| VL-RouterBench | Rank Score0.932 | 26 | 2mo ago | ||
| Model Routing Suite MathQA, LogiQA, MedQA, PIQA, TruthQA, MMLU, GSM8k, GPQA, ASDiv, SoQA | KMeans | Overall Accuracy66.2 | 18 | 4mo ago | |
| Dataset A | Accuracy (Selected Answer)83.25 | 13 | 1mo ago | ||
| NB-Curated OOD | Uniqueness35.4 | 11 | 3mo ago | ||
| NB-WildChat | Uniqueness Score42.6 | 11 | 3mo ago | ||
| OOD Set AIME Humanity's Last Exam SimpleQA OlympiadBench (test) | SCOPE | Avg. A50.8 | 11 | 4mo ago | |
| SCOPE-60K 5% split (test) | SCOPE | Avg. Accuracy75 | 11 | 4mo ago | |
| Model-Query Evaluation (test) | LOCUS | Routing Accuracy (%)64.7 | 9 | 4mo ago | |
| Global Routing Dataset Mixed Pool | SharedTrunkNet | P-AUCCC0.2323 | 7 | 4mo ago | |
| Global Routing Dataset Small Pool | SharedTrunkNet | P-AUCCC0.5472 | 7 | 4mo ago | |
| Global Routing Dataset Frontier Pool | SharedTrunkNet | P-AUCCC0.4377 | 7 | 4mo ago | |
| Small Pool | SharedTrunkNet | Oracle Accuracy92 | 6 | 4mo ago | |
| Frontier Pool | SharedTrunkNet | Oracle Accuracy89.3 | 6 | 4mo ago | |
| Mixed pool | SharedTrunkNet | Mean per-model AUC0.8817 | 6 | 4mo ago | |
| Small pool | SharedTrunkNet | Mean per-model AUC82.6 | 6 | 4mo ago | |
| Frontier pool | SharedTrunkNet | Mean per-model AUC0.856 | 6 | 4mo ago | |
| MathVista | P1 | Spearman Rank Correlation & Accuracy0.12 | 5 | 25d ago | |
| MathVerse | Spearman Rank Correlation (Acc)-0.01 | 5 | 25d ago | ||
| ChartQA | Spearman Rank Correlation Accuracy0 | 5 | 25d ago | ||
| AI Tutoring student queries | Alignment Rate (Tolerance 0.5)100 | 5 | 1mo ago | ||
| Model Routing (test) | JE-IRT | Routing Accuracy65.17 | 4 | 1mo ago | |
| EmbedLLM benchmark | ABLE | Router Accuracy67.6 | 4 | 1mo ago | |
| FRAMES (ID) | CPT (80%)60.92 | 4 | 4mo ago | ||
| MATH 500 (ID) | CPT (80%)53.79 | 4 | 4mo ago | ||
| AIME (ID) | CPT (80%)74.45 | 4 | 4mo ago |
