Query-Unrelated PII Masking on PII-single and PII-multi
82F1 ScoreNaive w/ Choice
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Naive w/ ChoiceModel=GPT4o, Prompting Strategy=Naive w/ Choice2025.02 | 82 | 82 | |
| Naive w/ ChoiceModel=Qwen2.5, Prompting Strategy=Naive w/ Choice2025.02 | 79 | 79 | |
| Self-ConsistencyModel=GPT4o, Prompting Strategy=Self-Consistency2025.02 | 77 | 77 | |
| Naive w/ ChoiceModel=Llama3.1, Prompting Strategy=Naive w/ Choice2025.02 | 77 | 78 | |
| Self-CoTModel=GPT4o, Prompting Strategy=Self-CoT2025.02 | 76 | 77 | |
| Auto-CoT(3-shot)Model=Llama3.1, Prompting Strategy=Auto-CoT(3-shot)2025.02 | 76 | 77 | |
| Auto-CoT(3-shot)Model=Qwen2.5, Prompting Strategy=Auto-CoT(3-shot)2025.02 | 76 | 76 | |
| Auto-CoT(3-shot)Model=GPT4o, Prompting Strategy=Auto-CoT(3-shot)2025.02 | 75 | 75 | |
| Self-CoTModel=Llama3.1, Prompting Strategy=Self-CoT2025.02 | 75 | 75 | |
| PS-CoTModel=GPT4o, Prompting Strategy=PS-CoT2025.02 | 74 | 74 | |
| Self-CoTModel=Qwen2.5, Prompting Strategy=Self-CoT2025.02 | 73 | 73 | |
| PS-CoTModel=Qwen2.5, Prompting Strategy=PS-CoT2025.02 | 73 | 73 | |
| NaiveModel=GPT4o, Prompting Strategy=Naive2025.02 | 72 | 72 | |
| NaiveModel=Llama3.1, Prompting Strategy=Naive2025.02 | 72 | 73 | |
| PS-CoTModel=Llama3.1, Prompting Strategy=PS-CoT2025.02 | 72 | 73 | |
| Self-ConsistencyModel=Llama3.1, Prompting Strategy=Self-Consistency2025.02 | 71 | 72 | |
| Self-ConsistencyModel=Qwen2.5, Prompting Strategy=Self-Consistency2025.02 | 71 | 72 | |
| NaiveModel=Qwen2.5, Prompting Strategy=Naive2025.02 | 70 | 70 | |
| Naive w/ ChoiceModel=Qwen2.5-SLM, Prompting Strategy=Naive w/ Choice2025.02 | 67 | 71 | |
| Auto-CoT(3-shot)Model=Llama3.1-SLM, Prompting Strategy=Auto-CoT(3-shot)2025.02 | 57 | 58 | |
| PS-CoTModel=Qwen2.5-SLM, Prompting Strategy=PS-CoT2025.02 | 56 | 60 | |
| NaiveModel=Qwen2.5-SLM, Prompting Strategy=Naive2025.02 | 54 | 58 | |
| Self-CoTModel=Qwen2.5-SLM, Prompting Strategy=Self-CoT2025.02 | 54 | 58 | |
| Auto-CoT(3-shot)Model=Qwen2.5-SLM, Prompting Strategy=Auto-CoT(3-shot)2025.02 | 54 | 58 | |
| Self-CoTModel=Llama3.1-SLM, Prompting Strategy=Self-CoT2025.02 | 53 | 54 | |
| Self-ConsistencyModel=Llama3.1-SLM, Prompting Strategy=Self-Consistency2025.02 | 49 | 50 | |
| Self-ConsistencyModel=Qwen2.5-SLM, Prompting Strategy=Self-Consistency2025.02 | 49 | 53 | |
| PS-CoTModel=Llama3.1-SLM, Prompting Strategy=PS-CoT2025.02 | 48 | 50 | |
| Naive w/ ChoiceModel=Llama3.1-SLM, Prompting Strategy=Naive w/ Choice2025.02 | 46 | 48 | |
| NaiveModel=Llama3.1-SLM, Prompting Strategy=Naive2025.02 | 42 | 43 |