List your company

Regional knowledge and conversational settings on CARE

51.26Average ScoreLRPO
14.91224.348533.78543.2215May 25, 2026

Evaluation results

MethodLinks
LRPOBackbone=Gemma3-4b-it2026.0551.26
GRPOBackbone=Gemma3-4b-it2026.0551.24
DPOBackbone=Gemma3-4b-it2026.0550.47
MAPOBackbone=Gemma3-4b-it2026.0550.22
VanillaBackbone=Gemma3-4b-it2026.0549.85
LIDRBackbone=Gemma3-4b-it2026.0549.23
MPOBackbone=Gemma3-4b-it2026.0548.98
LRPOBackbone=Qwen2.5-1.5b-it2026.0534.6
GRPOBackbone=Qwen2.5-1.5b-it2026.0533.66
DPOBackbone=Qwen2.5-1.5b-it2026.0532.69
VanillaBackbone=Qwen2.5-1.5b-it2026.0532.2
MAPOBackbone=Qwen2.5-1.5b-it2026.0531.98
MPOBackbone=Qwen2.5-1.5b-it2026.0531.91
LIDRBackbone=Qwen2.5-1.5b-it2026.0530.07
GRPOBackbone=LLaMA3.2-1b-it2026.0520.15
LRPOBackbone=LLaMA3.2-1b-it2026.0518.62
MPOBackbone=LLaMA3.2-1b-it2026.0517.67
VanillaBackbone=LLaMA3.2-1b-it2026.0517.51
DPOBackbone=LLaMA3.2-1b-it2026.0516.95
MAPOBackbone=LLaMA3.2-1b-it2026.0516.84
LIDRBackbone=LLaMA3.2-1b-it2026.0516.31