Question Answering on BoolQ (Accuracy and Exit Position)
85.9AccuracyMinistral3 8B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Ministral3 8BModel Architecture=Ministral3 8B (34 layers), Evaluation Protocol=Backbone2026.04 | 85.9 | — | |
| River-LLMModel Architecture=Ministral3 8B (34 layers), Early-exit threshold (tau)=0.9, Evaluation Protocol=Adaptive2026.04 | 85.4 | 3.83 | |
| River-LLMModel Architecture=Ministral3 8B (34 layers), Early-exit threshold (tau)=0.8, Evaluation Protocol=Adaptive2026.04 | 85.2 | 2.04 | |
| Phi4-miniModel Architecture=Phi4-mini (32 layers), Evaluation Protocol=Backbone2026.04 | 84.3 | — | |
| River-LLMModel Architecture=Phi4-mini (32 layers), Early-exit threshold (tau)=0.9, Evaluation Protocol=Adaptive2026.04 | 84.3 | 4.93 | |
| River-LLMModel Architecture=Phi4-mini (32 layers), Early-exit threshold (tau)=0.8, Evaluation Protocol=Adaptive2026.04 | 84.1 | 2.35 |