In-Context Learning Aggregate Evaluation for Probes on ICL composite Probes
20Macro AccuracyBaseline
Evaluation Results
| Method | Links | |
|---|---|---|
| BaselineModel Scale=1B2025.09 | 20 | |
| InductionModel Scale=1B2025.09 | 15.2 | |
| BalancedModel Scale=1B2025.09 | 14.9 | |
| Anti-inductionModel Scale=1B2025.09 | 14.7 |