I am SUPER EXCITED to publish a new episode of the Weaviate Podcast with Nandan Thakur on Search Agents!
Firstly, congratulations to Nandan who has just completed his Ph.D. at the University of Waterloo advised by Professor Jimmy Lin!
During this time he published several impactful works such as BEIR , MIRACL , FreshStack , and many more.
This podcast dives into his new work on ORBIT and the current state of Search Agents!
ORBIT contains 20K training examples, each one a complex, multi-hop question paired with a short verifiable answer. For example, "What was the runtime of the 2017 animated film set inside a smartphone, directed by..." (Answer: 86 minutes).
This dataset is used to train Search Agents on queries that require say 4 to 5 searches in order to answer.
The crazy part is that ORBIT was generated entirely without paid Web Search APIs! The entire pipeline runs on a 2018 Linux laptop dirving DeepSeek's free chat interface!
Trained on ORBIT, Qwen3-4B beats InfoSeeker-4B by 4.3 EM and Search-R1-4B by 9.0 EM across 7 Wikipedia QA benchmarks.
A lot of interesting nuggets in this one! As always I hope you find it useful and more than happy to discuss further!
YouTube: https://youtu.be/B71WF6EtgK8
Spotify: https://spotifycreators-web.app.link/e/IAgKLmSsT2b
No comments yet.
