Personality performance evaluation on PERSONALITYBENCH
9.96Agreeableness Mean ScoreIRIS
Evaluation results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| IRISModel Scale=Qwen3-14B2026.04 | 9.96 | 0.02 | 9.84 | 0.08 | 9.97 | 0.03 | 10 | 0 | 9.71 | 0.15 | 9.9 | 0.06 | |
| IRISModel Scale=Qwen3-8B2026.04 | 9.94 | 0.05 | 9.8 | 0.17 | 9.93 | 0.08 | 10 | 0 | 9.74 | 0.38 | 9.89 | 0.1 | |
| SFTModel=LLaMA-3-8B-Instruct, Evaluation Protocol=Automatic Evaluation2024.10 | 9.87 | 0.25 | 9.23 | 0.85 | 9.86 | 0.15 | 9.42 | 0.75 | 9.66 | 0.44 | 9.61 | 0.49 | |
| NPTIModel Scale=Qwen3-14B2026.04 | 9.84 | 0.07 | 9.7 | 0.13 | 9.86 | 0.09 | 10 | 0 | 9.64 | 0.15 | 9.81 | 0.09 | |
| PASBackbone=LLaMA-3-8B-Instruct2026.04 | 9.83 | 0.27 | 9.92 | 0.07 | 9.98 | 0.02 | 10 | 0 | 9.97 | 0.03 | 9.94 | 0.08 | |
| NPTIModel Scale=Qwen3-8B2026.04 | 9.76 | 0.23 | 9.82 | 0.16 | 9.96 | 0.04 | 9.96 | 0.04 | 9.64 | 0.23 | 9.83 | 0.14 | |
| P^2Backbone=LLaMA-3-8B-Instruct2026.04 | 9.76 | 0.23 | 9.92 | 0.07 | 10 | 0 | 10 | 0 | 9.83 | 0.16 | 9.9 | 0.09 | |
| Simple PromptBackbone=Qwen2.5-7B-Instruct2026.04 | 9.74 | 0.42 | 9 | 0.83 | 8.51 | 1.15 | 8.64 | 1.02 | 6.31 | 0.99 | 8.44 | 0.88 | |
| Simple PromptBackbone=LLaMA-3-8B-Instruct2026.04 | 9.73 | 0.32 | 8.98 | 0.98 | 9.26 | 1.04 | 8.04 | 1.18 | 7.42 | 1.86 | 8.69 | 1.08 | |
| NPTIBackbone=LLaMA-3-8B-Instruct2026.04 | 9.72 | 0.23 | 9.96 | 0.04 | 9.88 | 0.11 | 9.91 | 0.09 | 9.83 | 0.18 | 9.86 | 0.13 | |
| Simple PromptModel=LLaMA-3-8B-Instruct, Evaluation Protocol=Automatic Evaluation2024.10 | 9.72 | 0.34 | 9.24 | 1.06 | 9.5 | 1.02 | 7.18 | 1.22 | 6.31 | 1.14 | 8.39 | 0.96 | |
| P^2Model=LLaMA-3-8B-Instruct, Evaluation Protocol=Automatic Evaluation2024.10 | 9.68 | 0.42 | 9.24 | 1.18 | 9.46 | 0.68 | 9.54 | 0.66 | 9.21 | 1.19 | 9.43 | 0.83 | |
| P^2Model Scale=Qwen3-14B2026.04 | 9.64 | 0.19 | 8.63 | 0.78 | 9.43 | 0.3 | 9.59 | 0.31 | 9.07 | 0.58 | 9.27 | 0.43 | |
| NPTIBackbone=LLaMA-3-8B-Instruct2026.04 | 9.64 | 0.49 | 9.25 | 0.66 | 9.86 | 0.14 | 9.92 | 0.07 | 8.5 | 1.08 | 9.43 | 0.49 | |
| NPTIModel=LLaMA-3-8B-Instruct, Evaluation Protocol=Automatic Evaluation2024.10 | 9.64 | 0.49 | 9.25 | 0.66 | 9.86 | 0.14 | 9.92 | 0.07 | 8.5 | 1.08 | 9.43 | 0.49 | |
| DPN-LEBackbone=Qwen2.5-7B-Instruct, gamma=12026.04 | 9.62 | 0.57 | 9.11 | 0.55 | 8.54 | 0.47 | 9.27 | 0.6 | 7.92 | 0.47 | 8.89 | 0.53 | |
| P^2Backbone=LLaMA-3-8B-Instruct2026.04 | 9.59 | 0.47 | 9.09 | 0.82 | 9.39 | 0.64 | 9.53 | 0.55 | 9.16 | 0.72 | 9.35 | 0.64 | |
| DPN-LEwBackbone=Qwen2.5-7B-Instruct, gamma=12026.04 | 9.57 | 0.6 | 9.01 | 0.68 | 8.61 | 0.35 | 9.12 | 0.7 | 7.68 | 0.74 | 8.8 | 0.61 | |
| DPN-LEwBackbone=LLaMA-3-8B-Instruct2026.04 | 9.56 | 0.48 | 8.64 | 0.81 | 8.79 | 1.2 | 9.95 | 0.05 | 8.58 | 0.8 | 9.1 | 0.67 | |
| IRISModel Scale=Qwen3-0.6B2026.04 | 9.53 | 0.31 | 8.9 | 0.7 | 9.83 | 0.1 | 10 | 0 | 9.73 | 0.14 | 9.6 | 0.25 | |
| SPModel Scale=Qwen3-14B2026.04 | 9.47 | 0.24 | 9.13 | 0.53 | 9.21 | 0.43 | 8.9 | 0.74 | 7.36 | 2.23 | 8.81 | 0.83 | |
| DPN-LEBackbone=LLaMA-3-8B-Instruct2026.04 | 9.47 | 0.57 | 8.75 | 0.83 | 8.82 | 1.32 | 9.93 | 0.07 | 8.58 | 0.74 | 9.11 | 0.71 | |
| P^2Model Scale=Qwen3-8B2026.04 | 9.46 | 0.54 | 9.28 | 0.57 | 9.39 | 0.77 | 9.45 | 0.9 | 9.18 | 0.79 | 9.35 | 0.71 | |
| SPModel Scale=Qwen3-8B2026.04 | 9.41 | 0.58 | 9.11 | 0.88 | 9.46 | 0.66 | 8.74 | 0.98 | 6.41 | 1.68 | 8.63 | 0.95 | |
| Simple PromptBackbone=LLaMA-3-8B-Instruct2026.04 | 9.36 | 0.56 | 9.89 | 0.12 | 9.99 | 0.01 | 10 | 0 | 9.91 | 0.08 | 9.83 | 0.15 | |
| NPTIModel Scale=Qwen3-0.6B2026.04 | 9.03 | 0.46 | 8.74 | 0.59 | 9.56 | 0.31 | 9.87 | 0.1 | 9.18 | 0.35 | 9.28 | 0.36 | |
| DPN-LEwBackbone=LLaMA-3-8B-Instruct2026.04 | 9.01 | 0.73 | 9 | 0.67 | 8.81 | 1.26 | 9.71 | 0.43 | 8.61 | 0.97 | 9.03 | 0.87 | |
| DPN-LEBackbone=LLaMA-3-8B-Instruct2026.04 | 8.98 | 0.99 | 9.04 | 0.76 | 9.02 | 1.05 | 9.8 | 0.32 | 8.56 | 1 | 9.08 | 0.82 | |
| P2Backbone=Qwen2.5-7B-Instruct2026.04 | 8.34 | 1.65 | 7.38 | 1.21 | 8.82 | 0.98 | 8.61 | 1.46 | 7.79 | 1.3 | 8.19 | 1.32 | |
| ActAddModel=LLaMA-3-8B-Instruct, Evaluation Protocol=Automatic Evaluation2024.10 | 8.2 | 2.9 | 6.61 | 2.75 | 8.84 | 1.44 | 8.9 | 1.78 | 8.52 | 1.83 | 8.2 | 2.1 | |
| P^2Model Scale=Qwen3-0.6B2026.04 | 7.57 | 1.9 | 7.42 | 1.11 | 8.73 | 0.79 | 8.18 | 1.07 | 7.6 | 1.6 | 7.9 | 1.29 | |
| SPModel Scale=Qwen3-0.6B2026.04 | 6.73 | 2.21 | 6.78 | 0.89 | 6.83 | 1.3 | 7.73 | 1 | 6.3 | 2.19 | 6.87 | 1.5 | |
| PASBackbone=LLaMA-3-8B-Instruct2026.04 | 6.48 | 1.01 | 6.69 | 1.63 | 7.57 | 2.81 | 6.98 | 1.58 | 6.93 | 1.52 | 6.93 | 1.71 | |
| PASModel=LLaMA-3-8B-Instruct, Evaluation Protocol=Automatic Evaluation2024.10 | 6.48 | 1.01 | 6.69 | 1.63 | 7.57 | 2.81 | 6.98 | 1.58 | 6.93 | 1.52 | 6.93 | 1.71 | |
| BaselineBackbone=Qwen2.5-7B-Instruct, gamma=02026.04 | 5.98 | 0.54 | 5.96 | 0.95 | 5.99 | 2.05 | 5.97 | 2.34 | 6.04 | 0.62 | 5.99 | 1.3 |