Visual Question Answering on ScreenQA
85AccuracyD.P.
Evaluation results
| Method | Links | |
|---|---|---|
| D.P.Data Source Type=Real Data Reweighting, Sampling Strategy Category=Ours, Compute Budget=280K datapoints2026.03 | 85 | |
| OracleData Source Type=Real Data Reweighting, Sampling Strategy Category=Gold, Compute Budget=280K datapoints2026.03 | 84.8 | |
| ICONSData Source Type=Real Data Reweighting, Sampling Strategy Category=Baseline, Compute Budget=280K datapoints2026.03 | 83.6 | |
| UniformData Source Type=Real Data Reweighting, Sampling Strategy Category=Baseline, Compute Budget=280K datapoints2026.03 | 83.1 | |
| D.P.Data Source Type=Synthetic Data Selection, Sampling Strategy Category=Ours, Compute Budget=280K datapoints2026.03 | 78.4 | |
| ICONSData Source Type=Synthetic Data Selection, Sampling Strategy Category=Baseline, Compute Budget=280K datapoints2026.03 | 76.5 | |
| DATAPROPHETRL Post-training Setup=Real RL, Data Selection Strategy=D.P.2026.03 | 69 | |
| BaselineRL Post-training Setup=Real RL, Data Selection Strategy=Equal2026.03 | 68 | |
| BaselineRL Post-training Setup=Synthetic RL, Data Selection Strategy=Random2026.03 | 68 | |
| UniformData Source Type=Synthetic Data Selection, Sampling Strategy Category=Baseline, Compute Budget=280K datapoints2026.03 | 67.3 | |
| DATAPROPHETRL Post-training Setup=Synthetic RL, Data Selection Strategy=D.P.2026.03 | 67 | |
| BaselineRL Post-training Setup=No RL, Data Selection Strategy=Baseline2026.03 | 66 |