Neura Intelligence Index
Category Comparison
Benchmark-by-Benchmark
reasoning
| Benchmark | Llama 3.1 405B | Nova 2 Pro |
|---|---|---|
| GPQA Diamond | 51.1 | 81.4 |
| BigBench-Hard | 85.4 | — |
| ARC-Challenge | 96.9 | — |
general
| Benchmark | Llama 3.1 405B | Nova 2 Pro |
|---|---|---|
| MMLU-Pro | 73.3 | — |
| Chatbot Arena Elo | 1244 | — |
| IFEval | 86.4 | — |
language
| Benchmark | Llama 3.1 405B | Nova 2 Pro |
|---|---|---|
| WinoGrande | 87.6 | — |
| HellaSwag | 96 | — |
coding
| Benchmark | Llama 3.1 405B | Nova 2 Pro |
|---|---|---|
| HumanEval | 89 | — |
| SWE-bench Verified | — | 70 |
| TerminalBench | — | 41.3 |
safety
| Benchmark | Llama 3.1 405B | Nova 2 Pro |
|---|---|---|
| TruthfulQA | 65.3 | — |
multimodal
| Benchmark | Llama 3.1 405B | Nova 2 Pro |
|---|---|---|
| MMMU-Pro | — | 63.5 |
Category Winners
coding
69.3
math
Nova 2 Pro73.7
reasoning
Nova 2 Pro75.2
general
67.9
language
87.7
multimodal
Nova 2 Pro37.0
safety
59.8