/AI11d ago

Tülu researcher Rulin Shao releases DR Tulu, an open-source deep research model trained using reinforcement learning with evolving rubrics

The updated paper secured an ICML 2026 oral presentation.

039583.9K
#329Akari Asai
Rulin Shao
Original post
Rulin Shao@RulinShao#1228inAI

DR Tulu is now accepted for an oral presentation at #ICML2026 🙏

Updated paper: https://arxiv.org/abs/2511.19399 📥We added more ablations including using Qwen3-8B as the rubric generator&judge, showing evolving rubrics work with a weak model too; spurious rewards sanity check, etc.

Live demo: https://www.dr-tulu.org/ Code&models: https://github.com/rlresearch/dr-tulu

Rulin Shao@RulinShao

Happy to share that DR Tulu has been accepted to ICML as a ✨Spotlight✨!

We believe that co-evolving the agent and its reward metric can lead to more capable intelligence.

DR Tulu is a team effort. Huge thanks and congrats to all my amazing collaborators and mentors!

12:49 PM · May 25, 2026 · 13.7K Views
Sentiment
Sentiment building, check back later.
Cluster Engagement
Posts from X
Most Activity
Most Activity
VIEWS3.9KBOOKMARKS8LIKES39RETWEETS5
Akari Asai@AkariAsai

DR Tulu has been selected for an oral presentation at ICML 2026 (0.7% of all submissions) 🥳 Check out our latest version, featuring additional ablations and a deeper analysis of RL with evolving rubrics for unverifiable open ended tasks!

Rulin Shao@RulinShao

DR Tulu is now accepted for an oral presentation at #ICML2026 🙏

Updated paper: https://arxiv.org/abs/2511.19399 📥We added more ablations including using Qwen3-8B as the rubric generator&judge, showing evolving rubrics work with a weak model too; spurious rewards sanity check, etc.

Live demo: https://www.dr-tulu.org/ Code&models: https://github.com/rlresearch/dr-tulu

20hViews 3.9KLikes 39Bookmarks 8