Helpful Response Evaluation on HelpSteer-2
0.03CV (Helpfulness)Gemini 2.5 Flash
Evaluation results
| Method | Links | ||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Gemini 2.5 FlashTemperature=0, API=OpenRouter2026.01 | 0.03 | 0.86 | 0.1 | 0.72 | 0.16 | 0.68 | 1.08 | 0.87 | 0.17 | 0.87 | — | — | — | — | — | — | — | — | — | — | — | — | |
| GPT-4oTemperature=0, API=OpenRouter2026.01 | 0.08 | 0.84 | 0.11 | 0.76 | 0.24 | 0.67 | 0.16 | 0.89 | 0.04 | 0.83 | — | — | — | — | — | — | — | — | — | — | — | — | |
| LLaMA-4-MaverickTemperature=0, API=OpenRouter2026.01 | 0.13 | 0.64 | 0.24 | 0.54 | 0.25 | 0.4 | 0.25 | 0.77 | 0.33 | 0.75 | — | — | — | — | — | — | — | — | — | — | — | — | |
| Qwen3-30B-A3B-instructTemperature=0, API=OpenRouter2026.01 | 0.27 | 0.72 | 0.3 | 0.68 | 0.21 | 0.58 | 0.39 | 0.88 | 0.2 | 0.76 | — | — | — | — | — | — | — | — | — | — | — | — | |
| LLaMA-4-ScoutTemperature=0, API=OpenRouter2026.01 | 0.28 | 0.66 | 0.28 | 0.54 | 0.31 | 0.49 | 0.31 | 0.85 | 0.45 | 0.83 | — | — | — | — | — | — | — | — | — | — | — | — | |
| GPT-4o-miniTemperature=0, API=OpenRouter2026.01 | 0.3 | 0.67 | 0.12 | 0.6 | 0.4 | 0.54 | 0.3 | 0.81 | 0.21 | 0.74 | — | — | — | — | — | — | — | — | — | — | — | — | |
| Qwen3-235B-A22B-instructTemperature=0, API=OpenRouter2026.01 | 0.37 | 0.78 | 0.25 | 0.76 | 0.78 | 0.6 | 0.32 | 0.89 | 0.74 | 0.76 | — | — | — | — | — | — | — | — | — | — | — | — | |
| BoRPBackbone=Qwen3-14B, Model Size=14B, Paradigm=Probing2026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.695 | 0.696 | |
| Gemini 2.5 Flashvariant=Flash2026.01 | — | — | — | — | — | — | — | — | — | — | 1.8 | 0.33 | 1.46 | 0.28 | 1.1 | 0.25 | 1.26 | 0.34 | 1.02 | 0.37 | — | — | |
| GenBackbone=Qwen3-14B, Paradigm=Generative2026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.326 | 0.451 | |
| GenBackbone=Qwen3-Max, Paradigm=Generative2026.01 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 0.153 | 0.622 | |
| GPT-4o2026.01 | — | — | — | — | — | — | — | — | — | — | 1.83 | 0.33 | 1.55 | 0.29 | 1.29 | 0.23 | 1.16 | 0.3 | 1.21 | 0.45 | — | — | |
| GPT-4o-mini2026.01 | — | — | — | — | — | — | — | — | — | — | 1.76 | 0.32 | 1.77 | 0.32 | 1.34 | 0.15 | 1.5 | 0.32 | 1.46 | 0.17 | — | — | |
| Llama-4-Maverickvariant=Maverick2026.01 | — | — | — | — | — | — | — | — | — | — | 1.86 | 0.3 | 1.76 | 0.34 | 1.24 | 0.29 | 1.32 | 0.28 | 1.4 | 0.22 | — | — | |
| Llama-4-Scoutvariant=Scout2026.01 | — | — | — | — | — | — | — | — | — | — | 1.65 | 0.31 | 1.75 | 0.34 | 1.2 | 0.21 | 1.4 | 0.3 | 1.33 | 0.23 | — | — | |
| Qwen3-235bparameters=235B2026.01 | — | — | — | — | — | — | — | — | — | — | 1.58 | 0.33 | 1.6 | 0.3 | 1.48 | 0.15 | 1.3 | 0.36 | 1.18 | 0.52 | — | — | |
| Qwen3-30bparameters=30B2026.01 | — | — | — | — | — | — | — | — | — | — | 1.61 | 0.31 | 1.52 | 0.29 | 1.03 | 0.16 | 1.1 | 0.46 | 0.99 | 0.53 | — | — |