Alignment on IFEval strict prompt
90.2pass@1Nemotron Cascade-8B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Nemotron Cascade-8BParameters=8B, Thinking Mode=best of thinking/non-thinking2025.12 | 90.2 | — | |
| Gemini-2.5 Flash-ThinkingThinking Mode=true2025.12 | 89.8 | — | |
| SnapMLABackbone=LongCat-Flash-thinking, Precision=FP82026.02 | 87.8 | — | |
| SnapMLABackbone=DeepSeek-V3.1, Precision=FP82026.02 | 87.25 | — | |
| FlashMLABackbone=LongCat-Flash-thinking, Precision=BF162026.02 | 86.9 | — | |
| FlashMLABackbone=DeepSeek-V3.1, Precision=BF162026.02 | 86.32 | — | |
| Qwen3-14BModel Scale=14B, NGM Configuration=False, Decoding Settings=identical decoding settings2026.05 | 86.14 | — | |
| Nemotron-Nano 9B-v2Parameters=9B-v22025.12 | 86.1 | — | |
| Qwen3 14BParameters=14B2025.12 | 85.4 | — | |
| Qwen3 8BParameters=8B2025.12 | 85 | — | |
| Qwen3-8Bno think=true2026.02 | 84.29 | — | |
| DeepSeek-R1 0528 671BParameters=671B, Thinking Mode=true2025.12 | 84.1 | — | |
| Qwen3-8BModel Scale=8B, NGM Configuration=False, Decoding Settings=identical decoding settings2026.05 | 84.1 | — | |
| Qwen3-14B + NGMModel Scale=14B, NGM Configuration=True, Decoding Settings=identical decoding settings2026.05 | 83.92 | — | |
| Qwen3-8B + NGMModel Scale=8B, NGM Configuration=True, Decoding Settings=identical decoding settings2026.05 | 83.55 | — | |
| LLaDA2.1-minimode=Q Mode2026.02 | 83.18 | 1.25 | |
| Nemotron-Cascade 14B-ThinkingParameters=14B, Thinking Mode=best of thinking/non-thinking2025.12 | 81.9 | — | |
| LLaDA2.1-minimode=S Mode2026.02 | 81.33 | 1.83 | |
| LLaDA2.0-mini2026.02 | 80.78 | 1.24 | |
| Qwen3-4B + NGMModel Scale=4B, NGM Configuration=True, Decoding Settings=identical decoding settings2026.05 | 80.41 | — | |
| Qwen3-4BModel Scale=4B, NGM Configuration=False, Decoding Settings=identical decoding settings2026.05 | 80.22 | — | |
| Ling-mini-2.02026.02 | 76.16 | — | |
| Qwen3-1.7BModel Scale=1.7B, NGM Configuration=False, Decoding Settings=identical decoding settings2026.05 | 68.95 | — | |
| Qwen3-1.7B + NGMModel Scale=1.7B, NGM Configuration=True, Decoding Settings=identical decoding settings2026.05 | 67.47 | — | |
| Qwen3-0.6BModel Scale=0.6B, NGM Configuration=False, Decoding Settings=identical decoding settings2026.05 | 58.78 | — | |
| Qwen3-0.6B + NGMModel Scale=0.6B, NGM Configuration=True, Decoding Settings=identical decoding settings2026.05 | 57.86 | — |