Twitxy Saved

Hansi Zeng

@HansiZeng

CS PhD @ UMass Amherst CIIR | Prev Intern @GoogleDeepMind, @Amazon
hansizeng.github.io Joined March 2019
479 Following 353 Followers 59 Posts

Latest posts 5

Hansi Zeng @HansiZeng
Yeah! Excited to share that our paper has been accepted to COLM 2026! 😄
Hansi Zeng @HansiZeng Apr 27
What's bottlenecking your search agent?

Probably not the agent itself.

We show that jointly training the search agent and the ranker (retrieval) with RL improves performance by up to +10% relative F1 over Search-R1 across 7 QA benchmarks.

📄 arxiv.org/abs/2604.17555
View quoted post on X
0 replies 0 reposts 14 likes 814 views View on X
Hansi Zeng @HansiZeng
Can LLMs learn to search and reason interleavedly for complex QA tasks without predefined rules? Yes! 🚀 Check out Search-R1, our framework that leverages the DeepSeek-R1-style RL approach to achieve this.
Bowen Jin Verified account @BowenJin13 Mar 13, 2025
🚀 Excited to announce that our paper 𝐒𝐞𝐚𝐫𝐜𝐡-𝐑𝟏
is now live! 📄

We introduce an RL framework (an extension of 𝐃𝐞𝐞𝐩𝐬𝐞𝐞𝐤-𝐑𝟏) for training reasoning-and-retrieval interleaved LLMs. We’re also open-sourcing all resources—models, data, and more!

📜 Paper: arxiv.org/pdf/2503.09516
🛠️ Resources: huggingface.co/collections/Pe…

Check it out & let us know your thoughts! 🚀 #R1 #reasoning #AI #NLP #Search
View quoted post on X
Hansi Zeng @HansiZeng
Excited to be in DC for SIGIR! My first time for an in-person conference!

I'll present my paper, "Planning-Ahead in Generative Retrieval" tomorrow at 3:05 pm (Monday). Don't miss the M2.1 GenIR and future of LLMs for search session.

Looking forward to chatting with IR folks!