Datasets:

WaitHZ
/

SSPO-data

Tasks:

Question Answering

Text Generation

Modalities:

Formats:

Languages:

Size:

Tags:

reinforcement-learning

Libraries:

License:

Dataset card Data Studio Files Files and versions

WaitHZ commited on 4 days ago

Commit

a2db3e8

·

verified ·

1 Parent(s): a439d37

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -22,7 +22,7 @@ SSPO addresses this problem through two designs:
 Experiments with Qwen3-8B on BrowseComp, GAIA, and FRAMES show that SSPO consistently outperforms GRPO and matches or surpasses GRPO trained with twice as many gradient steps, while adding only about 5% training overhead from teacher scoring.
 <div align="center">
-  <img src="./assets/overview.png" width="100%">
 </div>
 ## 📚Dataset

 Experiments with Qwen3-8B on BrowseComp, GAIA, and FRAMES show that SSPO consistently outperforms GRPO and matches or surpasses GRPO trained with twice as many gradient steps, while adding only about 5% training overhead from teacher scoring.
 <div align="center">
+  <img src="./overview.png" width="100%">
 </div>
 ## 📚Dataset