Graduate-level Science Reasoning on GPQA (Score)
59.1ScoreQwen3-8B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3-8BParameters=8B2025.10 | 59.1 | |
| Qwen3-4BParameters=4B2025.10 | 54.5 | |
| Ouro-2.6B-Thinking-R4Parameters=2.6B, Configuration=Thinking-R42025.10 | 52.7 | |
| Deepseek-Distill-Qwen-7BParameters=7B2025.10 | 51 | |
| Ouro-1.4B-Thinking-R4Parameters=1.4B, Configuration=Thinking-R42025.10 | 45.5 | |
| Qwen3-1.7BParameters=1.7B2025.10 | 34 | |
| Deepseek-Distill-Qwen-1.5BParameters=1.5B2025.10 | 33.2 | |
| Nemotron-CC_ASI+Pre-training Curation Strategy=AI Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 27.1 | |
| DCLMPre-training Curation Strategy=Human Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 25.67 | |
| Fineweb-EduPre-training Curation Strategy=Human Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 24.51 | |
| Nemotron-CCPre-training Curation Strategy=AI Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 24.37 | |
| Nemotron-CC_ASIPre-training Curation Strategy=AI Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 23.93 | |
| Ultra-FinewebPre-training Curation Strategy=Human Discovered, Model Parameters=3B, Training Token Budget=500B tokens2026.03 | 23.04 | |
| GPT-o1Model Category=Reasoning Models, Assessed Version Date=2024-092025.11 | 0.78 | |
| DeepSeek–R1Model Category=Reasoning Models, Assessed Version Date=2025-052025.11 | 0.72 | |
| GPT-4oModel Category=Closed-Source Models (non-reasoning), Assessed Version Date=2025-032025.11 | 0.54 | |
| GPT-4o-MiniModel Category=Closed-Source Models (non-reasoning), Assessed Version Date=2024-112025.11 | 0.4 | |
| GPT-3.5-TurboModel Category=Closed-Source Models (non-reasoning), Assessed Version Date=2023-112025.11 | 0.31 | |
| Qwen2.5-3B-InstructModel Category=Open-Source Models (non-reasoning)2025.11 | 0.3 | |
| Falcon-7B-InstructModel Category=Open-Source Models (non-reasoning)2025.11 | 0.25 | |
| LLaMA2-13B-ChatModel Category=Open-Source Models (non-reasoning)2025.11 | 0.23 |