Open-ended Generation on CapArena
11.83Average ScoreGreedy
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GreedyBackbone=Qwen2.5-VL (3B/7B)2026.04 | 11.83 | 148.75 | |
| Min-pTemperature (τ)=1.0, Backbone=Qwen2.5-VL (3B/7B)2026.04 | 5.83 | 153.77 | |
| Top-kTemperature (τ)=1.0, Backbone=Qwen2.5-VL (3B/7B)2026.04 | 5.17 | 158.33 | |
| Top-p (nucleus)Temperature (τ)=1.0, Backbone=Qwen2.5-VL (3B/7B)2026.04 | 2.17 | 155.06 | |
| η-SamplingTemperature (τ)=1.0, Backbone=Qwen2.5-VL (3B/7B)2026.04 | 0.83 | 160.55 | |
| ε-SamplingTemperature (τ)=1.0, Backbone=Qwen2.5-VL (3B/7B)2026.04 | 0.67 | 156.53 | |
| Temp’ OnlyTemperature (τ)=1.0, Backbone=Qwen2.5-VL (3B/7B)2026.04 | -7.17 | 156.34 |