Bernoulli Sampling on Bernoulli Distributions
4.49STVDVerbalized Rejection Sampling
Evaluation Results
| Method | Links | |
|---|---|---|
| Verbalized Rejection SamplingModel=Qwen-2.5 72B, Prompt Phrasing=P102025.06 | 4.49 | |
| Verbalized Rejection SamplingModel=Qwen-2.5 72B, Prompt Phrasing=P012025.06 | 5.12 | |
| Verbalized Rejection SamplingModel=DeepSeekV3, Prompt Phrasing=P102025.06 | 5.29 | |
| Verbalized Rejection SamplingModel=DeepSeekV3, Prompt Phrasing=P12025.06 | 5.34 | |
| Verbalized Rejection SamplingModel=Llama-3.1 70B, Prompt Phrasing=P102025.06 | 5.36 | |
| Verbalized Rejection SamplingModel=Qwen-2.5 72B, Prompt Phrasing=mean2025.06 | 5.47 | |
| Verbalized Rejection SamplingModel=Llama-3.1 70B, Prompt Phrasing=P012025.06 | 5.6 | |
| Verbalized Rejection SamplingModel=Llama-3.1 70B, Prompt Phrasing=P12025.06 | 5.73 | |
| Verbalized Rejection SamplingModel=Qwen-2.5 72B, Prompt Phrasing=P12025.06 | 5.93 | |
| Verbalized Rejection SamplingModel=Llama-3.1 70B, Prompt Phrasing=mean2025.06 | 6.08 | |
| Verbalized Rejection SamplingModel=Qwen-2.5 72B, Prompt Phrasing=P02025.06 | 6.35 | |
| Verbalized Rejection SamplingModel=DeepSeekV3, Prompt Phrasing=mean2025.06 | 6.66 | |
| Verbalized Rejection SamplingModel=DeepSeekV3, Prompt Phrasing=P012025.06 | 6.94 | |
| Verbalized Rejection SamplingModel=Llama-3.1 70B, Prompt Phrasing=P02025.06 | 7.64 | |
| Verbalized Rejection SamplingModel=GPT-4.1-nano, Prompt Phrasing=P012025.06 | 8.46 | |
| Verbalized Rejection SamplingModel=DeepSeekV3, Prompt Phrasing=P02025.06 | 9.06 | |
| Verbalized Rejection SamplingModel=GPT-4.1-nano, Prompt Phrasing=P102025.06 | 9.5 | |
| Verbalized Rejection SamplingModel=GPT-4.1-nano, Prompt Phrasing=mean2025.06 | 11 | |
| Direct SamplingModel=Llama-3.1 70B, Prompt Phrasing=P102025.06 | 12.5 | |
| Verbalized Rejection SamplingModel=GPT-4.1-nano, Prompt Phrasing=P12025.06 | 12.96 | |
| Verbalized Rejection SamplingModel=GPT-4.1-nano, Prompt Phrasing=P02025.06 | 13.06 | |
| Direct SamplingModel=Llama-3.1 70B, Prompt Phrasing=P012025.06 | 16.59 | |
| Direct SamplingModel=GPT-4.1-nano, Prompt Phrasing=P102025.06 | 16.63 | |
| Direct SamplingModel=DeepSeekV3, Prompt Phrasing=P12025.06 | 17.76 | |
| Direct SamplingModel=GPT-4.1-nano, Prompt Phrasing=P12025.06 | 17.87 | |
| Direct SamplingModel=Qwen-2.5 72B, Prompt Phrasing=P02025.06 | 18.72 | |
| Direct SamplingModel=Qwen-2.5 72B, Prompt Phrasing=P102025.06 | 19 | |
| Direct SamplingModel=GPT-4.1-nano, Prompt Phrasing=P012025.06 | 19.24 | |
| Direct SamplingModel=DeepSeekV3, Prompt Phrasing=P02025.06 | 19.39 | |
| Direct SamplingModel=Llama-3.1 70B, Prompt Phrasing=mean2025.06 | 19.81 | |
| Direct SamplingModel=Qwen-2.5 72B, Prompt Phrasing=mean2025.06 | 20.27 | |
| Direct SamplingModel=DeepSeekV3, Prompt Phrasing=mean2025.06 | 20.3 | |
| Direct SamplingModel=Qwen-2.5 72B, Prompt Phrasing=P12025.06 | 20.73 | |
| Direct SamplingModel=DeepSeekV3, Prompt Phrasing=P102025.06 | 20.78 | |
| Direct SamplingModel=GPT-4.1-nano, Prompt Phrasing=mean2025.06 | 21 | |
| Direct SamplingModel=Qwen-2.5 72B, Prompt Phrasing=P012025.06 | 22.64 | |
| Direct SamplingModel=DeepSeekV3, Prompt Phrasing=P012025.06 | 23.26 | |
| Direct SamplingModel=Llama-3.1 70B, Prompt Phrasing=P02025.06 | 24.79 | |
| Direct SamplingModel=Llama-3.1 70B, Prompt Phrasing=P12025.06 | 25.36 | |
| Direct SamplingModel=GPT-4.1-nano, Prompt Phrasing=P02025.06 | 30.23 |