List your company

Open-ended Evaluation on BQD

8.93p+ ScoreDPO
4.31245.51126.717.9088Apr 14, 2026

Evaluation results

MethodLinks
DPOBackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.048.938.3
PFTBackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.048.698.29
CPOBackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.048.658.2
SFTBackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.048.638.18
SFTBackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.048.58.26
DPOBackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.048.457.96
CPOBackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.048.48.05
CAABackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.048.388.15
SFTBackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.048.378.16
CPOBackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.048.358.22
DPOBackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.048.168.28
CAABackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.048.038.18
BaseBackbone=LLAMA-3.1-8B-INSTRUCT, Training Protocol=Base, Preference Setting=-2026.047.778.23
PFTBackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.0467.97
SFTBackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.045.538.18
PFTBackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.045.498.39
CPOBackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.045.488.12
CAABackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.045.37.86
SFTBackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.045.298.27
CPOBackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.045.158.22
BaseBackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Base, Preference Setting=-2026.045.128.15
CAABackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.045.17.92
CPOBackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.045.088.3
SFTBackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.045.048.36
DPOBackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.044.988.28
DPOBackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.044.938.45
DPOBackbone=QWEN2.5-3B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.044.927.93
SFTBackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.044.97.57
CAABackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.044.827.55
DPOBackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.044.87.93
CPOBackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.044.787.62
SFTBackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.044.757.28
CPOBackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.044.77.35
BaseBackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Base, Preference Setting=-2026.044.637.61
DPOBackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p+2026.044.627.84
CPOBackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.044.67.55
CAABackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.044.567.68
SFTBackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Single Preference, Preference Setting=p-2026.044.527.47
DPOBackbone=QWEN2.5-7B-INSTRUCT, Training Protocol=Pair Preference, Preference Setting=-2026.044.497.94