Share your thoughts, 1 month free Claude Pro on usSee more

Soft constrained reward optimization on XSum on Gemma-7B

0.147Average Soft Constrained Reward

MLP-CSR

Updated 19d ago

Evaluation Results

Method	Links
MLP-CSR 2026.05		0.147
MLP-CSR 2026.05		0.147
MLP-CSR 2026.05		0.145
CSR 2026.05		0.144
CSR 2026.05		0.143
CSR 2026.05		0.143
MLP-CSR 2026.05		0.143
CSR 2026.05		0.143
MLP-CSR 2026.05		0.142
MLP-CSR 2026.05		0.142
MLP-CSR 2026.05		0.142
CSR 2026.05		0.142
MLP-CSR 2026.05		0.142
CSR 2026.05		0.141
CSR 2026.05		0.141
MLP-CSR 2026.05		0.14
CSR 2026.05		0.14
MLP-CSR 2026.05		0.139
MLP-CSR 2026.05		0.139
CSR 2026.05		0.137
CSR 2026.05		0.134
MLP-CSR 2026.05		0.123
CSR 2026.05		0.122
CSR 2026.05		0.117
Uniform 2026.05		0.066
Uniform 2026.05		0.044
Uniform 2026.05		0.037
Uniform 2026.05		0.037
Uniform 2026.05		0.037
Uniform 2026.05		0.036
Uniform 2026.05		0.03
Uniform 2026.05		0.021
Uniform 2026.05		0.021
Uniform 2026.05		0.021
Uniform 2026.05		0.015
Uniform 2026.05		0