List your company

Open-ended Dialogue Generation on CMD (test)

19.4Rouge-LNSR
14.51215.78117.0518.319May 27, 2026

Evaluation results

MethodLinks
NSR2026.0519.479.22287.952.1
W-REINFORCEpositive sample weight (w+)=0.05, negative sample weight (w-)=12026.0519.477.82287.451.7
NSR2026.0519.479.22287.952.1
W-REINFORCEw+=0.05, w-=12026.0519.477.82287.451.7
W-REINFORCEpositive sample weight (w+)=1, negative sample weight (w-)=52026.0518.779.32186.251.3
EAPO2026.0518.779.821.386.451.6
W-REINFORCEw+=1, w-=52026.0518.779.32186.251.3
EAPO2026.0518.779.821.386.451.6
GRPO2026.0518.678.320.683.850.3
W-REINFORCEpositive sample weight (w+)=0.2, negative sample weight (w-)=12026.0518.675.421.687.650.8
GRPO2026.0518.678.320.683.850.3
W-REINFORCEw+=0.2, w-=12026.0518.675.421.687.650.8
W-REINFORCEpositive sample weight (w+)=0.1, negative sample weight (w-)=12026.0518.580.820.787.151.8
W-REINFORCEw+=0.1, w-=12026.0518.580.820.787.151.8
W-REINFORCEpositive sample weight (w+)=1, negative sample weight (w-)=12026.0518.478.720.785.250.8
W-REINFORCEw+=1, w-=12026.0518.478.720.785.250.8
W-REINFORCEpositive sample weight (w+)=5, negative sample weight (w-)=12026.0517.375.219.983.849
W-REINFORCEw+=5, w-=12026.0517.375.219.983.849
W-REINFORCEpositive sample weight (w+)=1, negative sample weight (w-)=0.22026.051776.319.585.249.5
W-REINFORCEw+=1, w-=0.22026.051776.319.585.249.5
W-REINFORCEpositive sample weight (w+)=1, negative sample weight (w-)=0.052026.051674.918.482.748
W-REINFORCEw+=1, w-=0.052026.051674.918.482.748
W-REINFORCEpositive sample weight (w+)=1, negative sample weight (w-)=0.12026.0515.872.418.380.646.7
W-REINFORCEw+=1, w-=0.12026.0515.872.418.380.646.7
PSR2026.0514.768.817.382.145.7
PSR2026.0514.768.817.382.145.7