| question-answering-on-natural-questions | DeepSeek LLM 67B Chat | #13 | EM: 47.0 |
| common-sense-reasoning-on-arc-easy | DeepSeek LLM 67B Chat | #16 | Accuracy: 81.6 |
| instruction-following-on-ifeval | DeepSeek LLM 67B Chat | #19 | Prompt-level loose-accuracy: 59.1 |
| multi-task-language-understanding-on-c-eval | DeepSeek LLM 67B Chat | #23 | Accuracy: 65.2 |
| question-answering-on-piqa | DeepSeek LLM 67B Chat | #29 | Accuracy: 82.6 |
| common-sense-reasoning-on-winogrande | DeepSeek LLM 67B Chat | #53 | Accuracy: 76.0 |
| code-generation-on-humaneval | DeepSeek LLM 67B Chat | #76 | Pass@1: 73.8 |
| code-generation-on-mbpp | DeepSeek LLM 67B Chat | #84 | Accuracy: 61.4 |
| sentence-completion-on-hellaswag | DeepSeek LLM 67B Chat | #92 | Accuracy: 75.7 |
| arithmetic-reasoning-on-gsm8k | DeepSeek LLM 67B Chat | #126 | Accuracy: 84.1 |
| math-word-problem-solving-on-math | DeepSeek LLM 67B Chat | #131 | Accuracy: 32.6 |