Performance Estimation on DICES
0MAEBQ-RPF Rounded
Evaluation Results
| Method | Links | |
|---|---|---|
| BQ-RPF RoundedBudget=1%, Target Model=Gemini 3 Pro, Selection Strategy=Active Selection, Integer Rounding=true2026.04 | 0 | |
| BQ-SFBudget=1%, Target Model=Gemini 3 Pro, Selection Strategy=Active Selection2026.04 | 0.008 | |
| BQ-RPFModel=Claude 3.7 Sonnet, Budget=1%, Selection=Active2026.04 | 0.011 | |
| BQ-RPFSelection Mode=Active Selection, Budget=1%, Target Model=GPT-52026.04 | 0.012 | |
| BQ-SF RoundedSelection Mode=Active Selection, Budget=1%, Target Model=GPT-52026.04 | 0.012 | |
| BQ-TPFModel=Gemma 3 12B, Budget=1%, Selection Strategy=Active Selection2026.04 | 0.022 | |
| BQ-SFSelection Mode=Active Selection, Budget=1%, Target Model=GPT-52026.04 | 0.023 | |
| BQ-RPFBudget=1%, Model=Gemini 2.5 Pro, Selection=Active Selection2026.04 | 0.026 | |
| BQ-SF RoundedModel=Claude 3.7 Sonnet, Budget=1%, Selection=Active2026.04 | 0.027 | |
| BQ-RPF RandSelection Mode=Random Selection, Budget=1%, Target Model=GPT-52026.04 | 0.028 | |
| BQ-RPFModel=Gemma 3 12B, Budget=1%, Selection Strategy=Active Selection2026.04 | 0.031 | |
| BQ-SFScenario=New Model (NM), Feature Type=Score Features (SF), Labeling Budget=1%2026.04 | 0.038 | |
| BQ-RPF RandBudget=1%, Target Model=Gemini 3 Pro, Selection Strategy=Random2026.04 | 0.038 | |
| BQ-SF RoundedBudget=1%, Target Model=Gemini 3 Pro, Selection Strategy=Active Selection, Integer Rounding=true2026.04 | 0.038 | |
| BQ-TPF RandSelection Mode=Random Selection, Budget=1%, Target Model=GPT-52026.04 | 0.038 | |
| BQ-TPFTarget Model=Claude 4.5 Opus, Budget=1%, Selection Strategy=Active Selection + BQ2026.04 | 0.039 | |
| BQ-SFBudget=1%, Model=Gemini 2.5 Pro, Selection=Active Selection2026.04 | 0.041 | |
| BQ-SF RoundedBudget=1%, Model=Gemini 2.5 Pro, Selection=Active Selection2026.04 | 0.044 | |
| BQ-RPF RoundedModel=Claude 3.7 Sonnet, Budget=1%, Selection=Active2026.04 | 0.045 | |
| BQ-RPF RoundedTarget Model=Claude 4.5 Opus, Budget=1%, Selection Strategy=Active Selection + BQ2026.04 | 0.049 | |
| BQ-RPFBudget=1%, Target Model=Gemini 3 Pro, Selection Strategy=Active Selection2026.04 | 0.05 | |
| BQ-RPF RandBudget=1%, Target Model=Gemini 3 Flash, Selection=Random Selection + BQ2026.04 | 0.051 | |
| LR+LUREBudget=1%, Target Model=Gemini 3 Pro, Selection Strategy=Random2026.04 | 0.053 | |
| BQ-SF RandBudget=1%, Target Model=Gemini 3 Pro, Selection Strategy=Random2026.04 | 0.053 | |
| LR+LURETarget Model=Claude 4.5 Opus, Budget=1%2026.04 | 0.055 | |
| BQ-SFScenario=New Model (NM), Feature Type=Score Features (SF), Rounding=True, Labeling Budget=1%2026.04 | 0.056 | |
| BQ-SF RandBudget=1%, Model=Gemini 2.5 Pro, Selection=Random Selection2026.04 | 0.056 | |
| Random SamplingBudget=1%, Target Model=GPT-3.5 Turbo2026.04 | 0.057 | |
| BQ-TPF RandBudget=1%, Target Model=GPT-3.5 Turbo, Selection Type=Random Selection2026.04 | 0.057 | |
| LR+LUREBudget=1%, Target Model=Gemini 3 Flash2026.04 | 0.057 | |
| BQ-TPFBudget=1%, Target Model=Gemini 3 Flash, Selection=Active Selection + BQ2026.04 | 0.059 | |
| BQ-RPF RandModel=Gemma 3 12B, Budget=1%, Selection Strategy=Random Selection2026.04 | 0.059 | |
| RF+ISBudget=1%, Target Model=GPT-3.5 Turbo2026.04 | 0.06 | |
| Random SamplingBudget=1%, Target Model=Gemini 3 Flash2026.04 | 0.06 | |
| BQ-TPFBudget=1%, Target Model=GPT-3.5 Turbo, Selection Type=Active Selection2026.04 | 0.061 | |
| LR+LUREBudget=1%, Target Model=GPT-3.5 Turbo2026.04 | 0.063 | |
| BQ-SFSelection Strategy=Random Selection, Feature Type=Score Features (SF), Labeling Budget=1%2026.04 | 0.064 | |
| BQ-SFSelection Strategy=Active Selection, Feature Type=Score Features (SF), Rounding=True, Labeling Budget=1%2026.04 | 0.064 | |
| BQ-RPF RandBudget=1%, Target Model=GPT-3.5 Turbo, Selection Type=Random Selection2026.04 | 0.065 | |
| LR+ISBudget=1%, Target Model=Gemini 3 Pro, Selection Strategy=Random2026.04 | 0.066 | |
| BQ-SFSelection Strategy=Active Selection, Feature Type=Score Features (SF), Labeling Budget=1%2026.04 | 0.067 | |
| LR+ISBudget=1%, Target Model=GPT-52026.04 | 0.067 | |
| BQ-RPF RoundedModel=Gemma 3 12B, Budget=1%, Selection Strategy=Active Selection2026.04 | 0.067 | |
| BQ-TPF RandBudget=1%, Target Model=Gemini 3 Pro, Selection Strategy=Random2026.04 | 0.068 | |
| BQ-RPF RandBudget=1%, Model=Gemini 2.5 Pro, Selection=Random Selection2026.04 | 0.068 | |
| LR+ISTarget Model=Claude 4.5 Opus, Budget=1%2026.04 | 0.07 | |
| BQ-TPF RoundedBudget=1%, Target Model=Gemini 3 Pro, Selection Strategy=Active Selection, Integer Rounding=true2026.04 | 0.07 | |
| BQ-SF RandModel=Gemma 3 12B, Budget=1%, Selection Strategy=Random Selection2026.04 | 0.07 | |
| RF+LUREBudget=1%, Target Model=GPT-3.5 Turbo2026.04 | 0.071 | |
| BQ-TPF RandModel=Claude 3.7 Sonnet, Budget=1%, Selection=Random2026.04 | 0.071 | |
| BQ-RPFTarget Model=Claude 4.5 Opus, Budget=1%, Selection Strategy=Active Selection + BQ2026.04 | 0.072 | |
| LR+ISBudget=1%, Target Model=GPT-3.5 Turbo2026.04 | 0.072 | |
| BQ-RPFBudget=1%, Target Model=GPT-3.5 Turbo, Selection Type=Active Selection2026.04 | 0.072 | |
| BQ-SF RandSelection Mode=Random Selection, Budget=1%, Target Model=GPT-52026.04 | 0.072 | |
| RF+LUREBudget=1%, Target Model=Gemini 3 Flash2026.04 | 0.072 | |
| RF+LUREModel=Gemma 3 12B, Budget=1%2026.04 | 0.072 | |
| BQ-SF RandModel=Claude 3.7 Sonnet, Budget=1%, Selection=Random2026.04 | 0.073 | |
| BQ-RPF RoundedBudget=1%, Model=Gemini 2.5 Pro, Selection=Active Selection2026.04 | 0.073 | |
| BQ-RPF RoundedSelection Mode=Active Selection, Budget=1%, Target Model=GPT-52026.04 | 0.073 | |
| BQ-SFModel=Gemma 3 12B, Budget=1%, Selection Strategy=Active Selection2026.04 | 0.073 | |
| BQ-TPFSelection Strategy=Random Selection, Feature Type=Tuned Prompt Features (TPF), Labeling Budget=1%2026.04 | 0.078 | |
| BQ-SFModel=Claude 3.7 Sonnet, Budget=1%, Selection=Active2026.04 | 0.078 | |
| RF+ISLabeling Budget=1%2026.04 | 0.079 | |
| BQ-RPF RandTarget Model=Claude 4.5 Opus, Budget=1%, Selection Strategy=Random Selection + BQ2026.04 | 0.079 | |
| RF+LURELabeling Budget=1%2026.04 | 0.08 | |
| RF+LUREBudget=1%, Target Model=Gemini 3 Pro, Selection Strategy=Random2026.04 | 0.08 | |
| LR+LUREModel=Gemma 3 12B, Budget=1%2026.04 | 0.081 | |
| LR+LUREModel=Claude 3.7 Sonnet, Budget=1%2026.04 | 0.083 | |
| LR+ISBudget=1%, Target Model=Gemini 3 Flash2026.04 | 0.083 | |
| RF+ISBudget=1%, Target Model=Gemini 3 Flash2026.04 | 0.085 | |
| LR+ISLabeling Budget=1%2026.04 | 0.086 | |
| BQ-RPF RoundedBudget=1%, Target Model=GPT-3.5 Turbo, Selection Type=Active Selection2026.04 | 0.087 | |
| LR+ISBudget=1%, Model=Gemini 2.5 Pro2026.04 | 0.087 | |
| LR+LUREBudget=1%, Model=Gemini 2.5 Pro2026.04 | 0.087 | |
| BQ-RPF RandModel=Claude 3.7 Sonnet, Budget=1%, Selection=Random2026.04 | 0.091 | |
| LR+LUREBudget=1%, Target Model=GPT-52026.04 | 0.093 | |
| LR+ISModel=Gemma 3 12B, Budget=1%2026.04 | 0.093 | |
| BQ-TPF RoundedBudget=1%, Target Model=GPT-3.5 Turbo, Selection Type=Active Selection2026.04 | 0.095 | |
| BQ-RPFBudget=1%, Target Model=Gemini 3 Flash, Selection=Active Selection + BQ2026.04 | 0.096 | |
| RF+LURETarget Model=Claude 4.5 Opus, Budget=1%2026.04 | 0.099 | |
| BQ-TPF RandBudget=1%, Model=Gemini 2.5 Pro, Selection=Random Selection2026.04 | 0.099 | |
| Random SamplingSelection Strategy=Random Selection, Labeling Budget=1%2026.04 | 0.1 | |
| BQ-TPF RoundedModel=Claude 3.7 Sonnet, Budget=1%, Selection=Active2026.04 | 0.1 | |
| RF+LUREBudget=1%, Model=Gemini 2.5 Pro2026.04 | 0.1 | |
| BQ-RPFSelection Strategy=Random Selection, Feature Type=Raw Prompt Features (RPF), Labeling Budget=1%2026.04 | 0.105 | |
| BQ-TPFScenario=New Model (NM), Feature Type=Tuned Prompt Features (TPF), Labeling Budget=1%2026.04 | 0.105 | |
| BQ-SF RoundedModel=Gemma 3 12B, Budget=1%, Selection Strategy=Active Selection2026.04 | 0.107 | |
| BQ-SF RandBudget=1%, Target Model=GPT-3.5 Turbo, Selection Type=Random Selection2026.04 | 0.108 | |
| BQ-SFBudget=1%, Target Model=GPT-3.5 Turbo, Selection Type=Active Selection2026.04 | 0.108 | |
| BQ-SF RoundedBudget=1%, Target Model=GPT-3.5 Turbo, Selection Type=Active Selection2026.04 | 0.108 | |
| RF+LUREModel=Claude 3.7 Sonnet, Budget=1%2026.04 | 0.111 | |
| BQ-TPFModel=Claude 3.7 Sonnet, Budget=1%, Selection=Active2026.04 | 0.112 | |
| LR+ISModel=Claude 3.7 Sonnet, Budget=1%2026.04 | 0.115 | |
| RF+ISTarget Model=Claude 4.5 Opus, Budget=1%2026.04 | 0.117 | |
| RF+ISModel=Gemma 3 12B, Budget=1%2026.04 | 0.117 | |
| BQ-TPF RandModel=Gemma 3 12B, Budget=1%, Selection Strategy=Random Selection2026.04 | 0.12 | |
| BQ-TPFScenario=New Model (NM), Feature Type=Tuned Prompt Features (TPF), Rounding=True, Labeling Budget=1%2026.04 | 0.121 | |
| BQ-SF RoundedBudget=1%, Target Model=Gemini 3 Flash, Selection=Active Selection + BQ2026.04 | 0.121 | |
| Random SamplingModel=Gemma 3 12B, Budget=1%2026.04 | 0.121 | |
| BQ-TPF RandTarget Model=Claude 4.5 Opus, Budget=1%, Selection Strategy=Random Selection + BQ2026.04 | 0.123 |