Iris Xu

[Iris Xu @iris_j_xu]

Very excited to share our paper on Bandit Optimization for Agent Discovery! I'm grateful for the amazing mentorship from @ZhangWeiHong9, and to work with such a great team for my first paper: @GuangtaoZ, @ZexueHe, @charlesjin_, @ParejaAldo, @dannygut123, and @gan_chuang.

[Zhang-Wei Hong @ZhangWeiHong9]

🚀 Single-agent coding assistants hit a wall on long-horizon tasks. Even frontier AI IDEs like Claude Code rely on manual sub-agent design—which is tedious and suboptimal.

What if we could discover the optimal agent hierarchy automatically?

Introducing BOAD: Bandit Optimization for Agent Design.

🔥 Outperforms GPT-4 & Claude-3.7-Sonnet (36B model) 🤖 Uses Multi-Armed Bandits for principled agent discovery
Stop guessing prompts. Start optimizing. 🧵👇

This work is led by @iris_j_xu during her internship at MIT-IBM Watson AI Lab.

Paper: https://t.co/U576P0979m
Code: https://t.co/R2FfcBoSKu

#LLM #RL #AIagents #CodingAgents #MultiArmedBandit