List your company

Question Answering on HotpotQA (Adaptive Metrics Ac/Y)

84.5Accuracy (alpha=0.10)J-Q8B
40.92452.23763.5574.863May 13, 2026

Evaluation results

MethodLinks
J-Q8BModel=Qwen3-8B2026.0584.570.969.14089.61.3
J-Q8BModel=Gemma-4-E4B2026.0583.972.371.352.787.44.6
J-G26BModel=Gemma-4-E4B2026.0583.871.971.253.585.120.6
R-L8BModel=Gemma-4-E4B2026.0583.472.871.956.286.217.9
J-Q30BModel=Gemma-4-E4B2026.0583.472.771.856.383.719.1
PplModel=Gemma-4-E4B2026.0583.471.769.355.185.413.6
R-Q8BModel=Gemma-4-E4B2026.0583.172.471.453.78213.3
Ppl-StdModel=Gemma-4-E4B2026.0583.171.868.854.780.714.9
MVModel=Gemma-4-E4B2026.0582.872.4—33.3—0
Ppl-StdModel=Qwen3-8B2026.0581.969.566.653.683.114.2
J-Q30BModel=Qwen3-8B2026.0581.87168.551.184.63
MVModel=Qwen3-8B2026.0581.670.8—36.7—0
R-L8BModel=Qwen3-8B2026.0581.670.46953.383.213.1
PplModel=Qwen3-8B2026.0581.669.666.953.682.910.5
R-Q8BModel=Qwen3-8B2026.0581.470.568.854.777.811.2
J-G26BModel=Qwen3-8B2026.0581.170.867.853.682.48.5
OracleModel=Gemma-4-E4B2026.0579.1—79.110079.1100
OracleModel=Qwen3-8B2026.0577.8—77.810077.8100
OracleModel=Qwen3-1.7B2026.0573.9—73.910073.9100
GreedyModel=Gemma-4-E4B2026.0570——10070100
MVModel=Qwen3-1.7B2026.0568.651.9—24.472.212.7
J-Q8BModel=Qwen3-1.7B2026.0568.652.245.226.671.213.3
J-Q30BModel=Qwen3-1.7B2026.0568.351.145.527.270.914.2
PplModel=Qwen3-1.7B2026.0568.148.540.830.272.815.5
GreedyModel=Qwen3-8B2026.0567.6——10067.6100
Ppl-StdModel=Qwen3-1.7B2026.0567.547.138.829.773.115
J-G26BModel=Qwen3-1.7B2026.0567.450.539.827.870.115.1
R-L8BModel=Qwen3-1.7B2026.0566.848.942.42768.214.7
R-Q8BModel=Qwen3-1.7B2026.0564.547.540.226.46814.2
GreedyModel=Qwen3-1.7B2026.0542.6——10042.6100