General Knowledge on CEval
90.4ScoreLongCat-Flash Chat
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| LongCat-Flash ChatEvaluation Mode=Chat2025.12 | 90.4 | — | |
| LongCat-Flash Exp-ChatEvaluation Mode=Chat2025.12 | 89.9 | — | |
| GLM 4.6Evaluation Mode=Chat2025.12 | 89.6 | — | |
| DeepSeek V3.2Evaluation Mode=Chat2025.12 | 89.6 | — | |
| T3STraining Dataset=BOBA-200, Teacher Model=DeepSeek-R1, Student Model=Qwen3-8B2026.01 | 84.4 | 0.07 | |
| T3STraining Dataset=S1K-200, Teacher Model=DeepSeek-R1, Student Model=Qwen3-8B2026.01 | 83.8 | 0 | |
| Qwen3-8B (Base)Student Model=Qwen3-8B2026.01 | 83.58 | — | |
| T3STraining Dataset=BOBA-200, Teacher Model=QWQ, Student Model=Qwen3-8B2026.01 | 83.58 | 0.04 | |
| T3STraining Dataset=S1K-200, Teacher Model=QWQ, Student Model=Qwen3-8B2026.01 | 83.51 | 0.4 | |
| SFTTraining Dataset=S1K-200, Teacher Model=QWQ, Student Model=Qwen3-8B2026.01 | 83.12 | -0.58 | |
| SFTTraining Dataset=S1K-200, Teacher Model=DeepSeek-R1, Student Model=Qwen3-8B2026.01 | 82.88 | -0.46 | |
| SFTTraining Dataset=BOBA-200, Teacher Model=DeepSeek-R1, Student Model=Qwen3-8B2026.01 | 82.84 | -0.75 | |
| SFTTraining Dataset=BOBA-200, Teacher Model=QWQ, Student Model=Qwen3-8B2026.01 | 82.17 | -1.26 | |
| Full AttentionSetting=CPT setting2026.04 | 73.66 | — | |
| Hybrid-KSASetting=CPT setting2026.04 | 72.66 | — | |
| Hybrid-SWASetting=CPT setting2026.04 | 72.36 | — | |
| KSASetting=CPT setting2026.04 | 72.14 | — | |
| Hybrid-SCASetting=CPT setting2026.04 | 71.66 | — | |
| Hybrid-LinearSetting=CPT setting2026.04 | 67.42 | — |