Text Summarization on XSum (ROUGE-2, Final Gap)
43.1ROUGE-2Pioneer Agent
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Pioneer AgentModel=Qwen3-8B, System Strategy=Agent, Checkpoint Stage=Best2026.04 | 43.1 | 36.6 | |
| Pioneer AgentModel=Qwen3-8B, System Strategy=Agent, Checkpoint Stage=C12026.04 | 38.9 | 36.6 | |
| Naive BaselineModel=Qwen3-8B, System Strategy=Naive, Checkpoint Stage=C12026.04 | 10.9 | 36.6 | |
| Naive BaselineModel=Qwen3-8B, System Strategy=Naive, Checkpoint Stage=C42026.04 | 6.5 | 36.6 |