Xuefeng Liu

Postdoc, @Stanford Medicine | Ph.D., @UChicago Computer Science

[AI for Drug Discovery and Development @AI4D3]

It’s a wrap! 🎉
Thanks to all speakers, organizers, and participants who made the AI4D3 @ NeurIPS2025 workshop a success.
Photos + highlights👇
🔗 https://t.co/PMXlrBhARw

[Michelle M. Li (李敏蕊) @DrMichelleMLi]

Our @NeurIPSConf workshop ✨AI Virtual Cells & Instruments✨ has a superstar lineup of roundtable discussions!

🛠️Building AI Virtual Cells & Instruments
💊Lessons Learned: AI-first Drug Discovery & Development

Join us on Dec 6 @ San Diego Convention Center Upper Level Rm 28A-E!

[AI for Drug Discovery and Development @AI4D3]

(4/6) The invited speakers are @biogerontology (Insilico Medicine), Tommi S. Jaakkola (MIT), Rick L. Stevens (UChicago & Argonne NL), @jinboxu_chicago (TTIC & Molecule Mind), @ericxing (MBZUAI, GenBio, & CMU), @MengdiWang10 (Princeton), and @LindaBGoodman (FaunaBio).

[AI for Drug Discovery and Development @AI4D3]

@biogerontology @jinboxu_chicago @ericxing @MengdiWang10 @LindaBGoodman (5/6) The organizers are @QuanquanGu (UCLA & ByteDance), @_michellemli (Harvard), @ChongLiuCS (UAlbany), @XuefengCS (UChicago), Abhishek Pandey (AbbVie), @tagasovska (Prescient Design - Genentech), and @marinkazitnik (Harvard).

[AI for Drug Discovery and Development @AI4D3]

(1/6)
We’re thrilled 🎉 to launch the #NeurIPS2025 Workshop on AI Virtual Cells and Instruments: A New Era in Drug Discovery & Development (AI4D3-2025) in San Diego, CA on December 6 or 7!🥳
🔗Workshop site: https://t.co/V20jrOeyNx

[AI for Drug Discovery and Development @AI4D3]

🔍 Call for Reviewers: AI4D3@NeurIPS 2025

We're seeking experts in AI & drug discovery to review submissions for our NeurIPS workshop.

🗓️ Deadline: May 30, 2025
�� Workshop: Dec 2025, San Diego
✅ Sign up: https://t.co/VDZIKa6i2I

#AI4D3 #NeurIPS2025 #DrugDiscovery #AI

[Xuefeng Liu @XuefengCS]

Form to register your attendance and (optionally) talk/poster presentation:
https://t.co/C7YyDgJfdq

[Xuefeng Liu @XuefengCS]

Join us for an exciting workshop about cutting-edge AI for drug discovery and development on April 4, 2025, at University of Chicago 🙌 #AI4D3

[AI for Drug Discovery and Development @AI4D3]

Join us at AI4D3-Midwest-2025 on April 4, 2025, at the University of Chicago!

🔥Cutting-edge AI for drug discovery & development
💡Top researchers & industry leaders
🤝Network & collaborate

FREE registration! Details: https://t.co/AFMIZY3rNM

[Xuefeng Liu @XuefengCS]

#ICML2024 I will present Entropy-Reinforced Planning with Large Language Models for Drug Discovery poster! Come and chat if you are interested!! Hall C 4-9 #2410 Thu 25 Jul 1:30 p.m. CEST — 3 p.m. CEST Paper: https://t.co/76CZ55z8Px

[Chong Liu @ChongLiuCS]

Call for reviewers of 2nd Workshop on AI for Drug Discovery and Development @AI4D3 at @NeurIPSConf '24! Feel free to sign up and share!
https://t.co/d62zbSe4LB

[Xuefeng Liu @XuefengCS]

website: https://t.co/m6w9nXuWPL

[Xuefeng Liu @XuefengCS]

It’s great to introduce <Blending IL and RL for Robust Policy Improvement> as a #SPOTLIGHT presentation at #ICLR2024 (joint work with @takuma_yoneda, Rick Stevens, @mattrwalter @yuxinch). Thanks to quite a few people who stopped by with interesting questions and follow-ups!

[Xuefeng Liu @XuefengCS]

Happening Now! Excited to deliver an oral presentation at #NeurIPS23 AI for Drug Discovery workshop (AI4D3).
📜"DrugImprover: Utilizing Reinforcement Learning for Multi-Objective Alignment in Drug Optimization"
📍Room #242
⏰ 1:20 p.m- 2:25 p.m. CST
🔗 https://t.co/A7CV7X9UeP

[Xuefeng Liu @XuefengCS]

If you are attending #ICML2023, come by our poster on Tuesday at 11am @ Exhibit Hall 1 to have a chat!
https://t.co/W4W9Ej0gcK

[Xuefeng Liu @XuefengCS]

How can an agent actively learn from multiple blackbox oracles by taking advantage of their complementary expertise to learn a better policy in a sample-efficient manner?

In our #ICML2023 paper, we introduce MAPS (Active Policy Selection) and MAPS-SE (+Active State Exploration)

[Xuefeng Liu @XuefengCS]

In particular, MAPS actively selects which of the oracles to imitate and improve their value function estimates, and MAPS-SE additionally leverages an active state exploration criterion to determine which state one should explore.

[Xuefeng Liu @XuefengCS]

Hello Twitter! #myfirstTweet