Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
safety
| Benchmark | Mixtral 8x22B | Phi 4 |
|---|---|---|
| TruthfulQA | 55.2 | — |
coding
| Benchmark | Mixtral 8x22B | Phi 4 |
|---|---|---|
| HumanEval | 79.3 | — |
language
| Benchmark | Mixtral 8x22B | Phi 4 |
|---|---|---|
| WinoGrande | 82.1 | — |
| HellaSwag | 88.5 | — |
reasoning
| Benchmark | Mixtral 8x22B | Phi 4 |
|---|---|---|
| ARC-Challenge | 91.3 | — |
| GPQA Diamond | — | 56.1 |
general
| Benchmark | Mixtral 8x22B | Phi 4 |
|---|---|---|
| MMLU-Pro | 56.8 | — |
| IFEval | 72.4 | — |
| Chatbot Arena Elo | 1146 | — |
| SimpleQA | — | 3 |
Category Winners
coding
26.8
math
16.4
reasoning
29.8
general
8.7
language
22.6
safety
8.2