Built with and by Teycir Ben Soltane•
How to Use•FAQ•GitHub•arXiv.org•
Share:
ArXivCSExplorer
☆☆Bookmarks🏆RSSHow to UseFAQ
Home/Authors/Songjun Tu

Songjun Tu

2 indexed papers

Recent (6 mo)
2
With code
0
Influential cites
0
Benchmarked
0

Publications per year

2
26

Top categories

NLP×2

Frequent co-authors

Yaocheng Zhang2×
Jiajun Chai2×
Yuqian Fu2×
Xiaohan Wang2×
Wei Lin2×
Guojun Yin2×

Research Timeline

2026
Are Full Rollouts Necessary for On-Policy Distillation?

This paper proposes two horizon-control strategies, Progressive OPD (POPD) and Truncated OPD (TOPD), demonstrating that full rollouts are often unnecessary for On-Policy Distillation, leading to significant improvements in training efficiency.

Are Full Rollouts Necessary for On-Policy Distillation?

This paper proposes two horizon-control strategies, Progressive OPD (POPD) and Truncated OPD (TOPD), demonstrating that full rollouts are often unnecessary for On-Policy Distillation, leading to significant improvements in training efficiency.

Highlighted terms show continued research focus across papers

Papers

cs.CLRecentMay 29, 2026

Are Full Rollouts Necessary for On-Policy Distillation?

Yaocheng Zhang, Jiajun Chai, Yuqian Fu, Songjun Tu +6 more

This paper proposes two horizon-control strategies, Progressive OPD (POPD) and Truncated OPD (TOPD), demonstrating that full rollouts are often unnecessary for On-Policy Distillation, leading to signi…

View →
cs.CLRecentMay 29, 2026

Are Full Rollouts Necessary for On-Policy Distillation?

Yaocheng Zhang, Jiajun Chai, Yuqian Fu, Songjun Tu +6 more

This paper proposes two horizon-control strategies, Progressive OPD (POPD) and Truncated OPD (TOPD), demonstrating that full rollouts are often unnecessary for On-Policy Distillation, leading to signi…

View →