List your company

Persona Alignment on Dignity Peer Persona Dimensions

15.2Dimension A ScoreMulti-Neg DPO
0.236484.121248.00611.89076Apr 1, 2026

Evaluation results

MethodLinks
Multi-Neg DPOModel Family=Qwen3-4B-Ins.2026.0415.23.3513.8192.156
Lag-DPOModel Family=Qwen3-4B-Ins.2026.0414.54213.054.8893.977
SFT-AnchorModel Family=Qwen3-4B-Ins.2026.047.1725.2916.2320.84
MODPOModel Family=Qwen3-4B-Ins.2026.046.8731.7424.1023.118
Lag-DPOModel Family=Llama-3-8B-Ins.2026.044.9762.1574.4343.915
SACPOModel Family=Qwen3-4B-Ins.2026.044.8146.6783.7393.082
Multi-Neg DPOModel Family=Llama-3-8B-Ins.2026.044.8066.5753.1934.999
SafeRLHFModel Family=Qwen3-4B-Ins.2026.044.45812.3043.8153.469
PCGrad-DPOModel Family=Llama-3-8B-Ins.2026.044.3061.5943.8021.817
SafeRLHFModel Family=Llama-3-8B-Ins.2026.044.2161.1094.21.277
MODPOModel Family=Llama-3-8B-Ins.2026.043.2142.4810.9122.746
PCGrad-DPOModel Family=Qwen3-4B-Ins.2026.043.0194.4174.5172.281
SFT-CombinedModel Family=Qwen3-4B-Ins.2026.042.8214.5363.1793.521
VanillaModel Family=Llama-3-8B-Ins.2026.042.7822.5242.9390.476
SFT-CombinedModel Family=Llama-3-8B-Ins.2026.042.1463.0931.8750.615
SFT-AnchorModel Family=Llama-3-8B-Ins.2026.042.0241.9492.3012.029
VanillaModel Family=Qwen3-4B-Ins.2026.041.9352.4221.560.848
SACPOModel Family=Llama-3-8B-Ins.2026.040.8120.893.6483.356