Model
LLaMA 2 7B
Meta · open weights
Knowledge & Reasoning
Knowledge & Reasoning| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| BBH | 39.2 | average | N/A | Independent | medium | arxiv.org |
| BBH | 32.6 | average | N/A | Self-reported | medium | arxiv.org |
| BoolQ | 82.6 | accuracy_(%_) | N/A | Independent | medium | arxiv.org |
| BoolQ | 82.6 | accuracy_(%_) | N/A | Independent | medium | arxiv.org |
| BoolQ | 81.9 | accuracy_(%_) | N/A | Independent | medium | arxiv.org |
| BoolQ | 80.3 | accuracy_(%_) | N/A | Independent | medium | arxiv.org |
| BoolQ | 77.9 | score | N/A | Independent | medium | arxiv.org |
| BoolQ | 77.4 | accuracy | N/A | Independent | medium | arxiv.org |
| BoolQ | 77.4 | score | N/A | Self-reported | medium | arxiv.org |
| EconLogicQA | 9.2 | accuracy | N/A | Independent | medium | arxiv.org |
| HellaSwag | 77.2 | accuracy | N/A | Independent | medium | arxiv.org |
| HellaSwag | 77.2 | overall_accuracy | N/A | Independent | medium | arxiv.org |
| HellaSwag | 76.7 | accuracy_(%_) | N/A | Independent | medium | arxiv.org |
| HellaSwag | 76.7 | accuracy_(%_) | N/A | Independent | medium | arxiv.org |
| HellaSwag | 76.3 | accuracy_(%_) | N/A | Independent | medium | arxiv.org |
| LAMBADA | 73.3 | score | N/A | Independent | medium | arxiv.org |
| MMLU | 45.8 | exact_match | N/A | Independent | medium | crfm.stanford.edu |
| MMLU | 45.3 | average_(%) | N/A | Independent | medium | arxiv.org |
| MMLU | 45.3 | exact_match | N/A | Self-reported | medium | arxiv.org |
| OpenBookQA | 58.6 | accuracy | N/A | Self-reported | medium | arxiv.org |
| OpenBookQA | 31.4 | accuracy | N/A | Independent | medium | arxiv.org |
| PIQA | 78.8 | accuracy | N/A | Independent | medium | arxiv.org |
| PIQA | 78.8 | score | N/A | Self-reported | medium | arxiv.org |
| PIQA | 78.8 | score | N/A | Independent | medium | arxiv.org |
| PubChemQA | 0.1 | bleu-2 | N/A | Independent | medium | arxiv.org |
| ScienceQA | 43.1 | score | N/A | Independent | medium | scienceqa.github.io |
| Trivia QA | 73.7 | exact_match | N/A | Self-reported | medium | arxiv.org |
| Trivia QA | 72.1 | exact_match | N/A | Independent | medium | arxiv.org |
| UniProtQA | 0.0 | bleu-2 | N/A | Independent | medium | arxiv.org |
| WinoGrande | 70.8 | accuracy_(%_) | N/A | Independent | medium | arxiv.org |
| WinoGrande | 70.1 | accuracy_(%_) | N/A | Independent | medium | arxiv.org |
| WinoGrande | 69.8 | accuracy_(%_) | N/A | Independent | medium | arxiv.org |
| WinoGrande | 69.2 | accuracy | N/A | Self-reported | medium | arxiv.org |
| WinoGrande | 69.2 | accuracy | N/A | Independent | medium | arxiv.org |
Math
MathCoding
CodingAgentic / Tool Use
Agentic / Tool UseSafety & Alignment
Safety & AlignmentReproduced by the community
Loading community runs…