Impressive paper showing how much the first retrieval step matters for deep research…
This is a AI post classified by Jev as AI agents (research), kept by the AI Radar because it carries real work, not commentary.
Impressive paper showing how much the first retrieval step matters for deep research agents. It helps to improve GPT-5.5 from 83.1% to 90.5% on BrowseComp-Plus with the same retriever and the same agent loop. It seems that the gain comes from the opening context. The authors propose Question's Gambit which runs once, before the agent starts searching. It splits the question into clues, turns each clue into complementary searches, pools the results, and reranks them. The agent then starts its loop with that ranked set already in context. The same change lifts GPT-5.4-mini from 68.1% to 79
Posted by elvis (320.8k followers) 18 h ago · 100 likes · 10.7k views · view the original post on X. Kept by the AI Radar as AI agents. Tools mentioned: academy.dair.ai.
More AI work like this
- Having @grok Build submit a PR for multiple profiles in Harness by @autonomous_labs so… — @MichaelGannotti
- BREAKING: SpaceXAI just released a major Grok Build update, with Grok 4.7 now available… — @cb_doge
- If your agents are making slop, don't give up or fight the agent with cleanup. Work on… — @kentcdodds
- Crazyy timeline....Opus 5.5 just dropped. — @Saboo_Shubham_
- Easily one of the most impressive use-cases for agentic AI that I've come across thus far. — @farzyness
- Opus-5.5 in the Hermes Agent model menu via the Nous Portal🤩 — @mr_r0b0t
- Univer is now Top 3 on GitHub Trending.🚀 — @LeonUniver
- And Harness from @autonomous_labs is now setup on my Omarchy Linux laptop and ready to… — @MichaelGannotti
Every post is read and classified by Jev (TypeSafe): what it is, which market it belongs to, and whether the link is a real tool. 42.4k posts from 5k X accounts over the last 14 days, 1.8k tools, 19 markets. Collected every 5 minutes, fully re-ranked every hour — last update 2026-09-22 17:33 UTC. Full method.