PostDoc at Stanford | PhD in Computer Science at NCSU l ML in Healthcare & Education
- Tweets 26
- Following 65
- Followers 34
- Likes 69
Ge Gao retweeted
Evaluating tool-use capabilities is expensive and domain-dependent. Once existing test data become saturated, expert humans need to design new tasks. Alternatively, LLMs can auto-generate challenging tasks from existing toolsets and task primitives through evolutionary search using multiple agents.
With @MichaelD1729, @natashajaques, @minsuk_chang.
Paper: arxiv.org/abs/2610.27307 for how to auto-generate tasks from existing benchmarks and use them for failure diagnosis
6
7
29
5,092
Ge Gao retweeted
Will be in SF for #CoLM2026! I’m presenting GIANTS and Librarian Agents (more to come soon 👀).
Interested in test-time adaptation, learning in open-ended environments, and self-improvement.
Also, would be on the job market this year. Feel free to reach out (dms open)!
🚨🚨New Paper: Generative Insight Anticipation from Scientific Literature
Human scientists achieve breakthroughs by "standing on the shoulders of giants," synthesizing profound insights from disparate sources. While LMs show promise in scientific discovery, they currently struggle to reliably generate ideas of true impact, diversity, and feasibility.
Introducing "insight anticipation," a novel task challenging LLMs to generate a downstream paper's core scientific contribution directly from its foundational prior works. We also introduce GIANTS-4B, an RL-trained model that significantly outperforms frontier models at this task! 🧵⬇️
3
4
56
2,965
Ge Gao retweeted
✨New work on off-policy evaluation with synthetic trajectories: PERRY (Policy Evaluation with Confidence Intervals using Auxiliary Data), accepted to TMLR!✨
TL;DR: we incorporate fully synthetic trajectories into off-policy evaluation (OPE) and develop two methods to learn valid confidence intervals over the policy value.
Paper: arxiv.org/abs/2507.20068
Authors: @Aishwarya_R_M , @jasonjmeng , @GeGaoCS , @JiankaiSun , @MacSchwager, @BeEngelhardt , @EmmaBrunskill
6
6
33
2,374
Ge Gao retweeted
Excited to release our latest paper, Maximizing Mutual Information Between Prompt and Response Improves LLM Performance With No Additional Data (in short, MIPO), which has been accepted to ICML 2026!
tl;dr Can we build a data augmentation method analogous to translation and rotation in computer vision for training LLMs without external verifiers or human supervision? We show that generating chosen responses from correct prompts and rejected responses from misspecified (random!) prompts can provide promising training signals.
w/ @AimeeLiS @natashajaques
paper: arxiv.org/abs/2603.19294
code: github.com/nam630/mutual_inf…
2
7
28
13,870
Ge Gao retweeted
Introducing EXPO-FT – Efficient, Reliable & Open-Source VLA Finetuning!
EXPO-FT unlocks π0.5 for challenging manipulation tasks:
Routing string lights & inserting the power connector to illuminate them
Striking pool ball into pocket
Inserting flower into wine bottle
(1/5)
13
31
240
79,966
Ge Gao retweeted
Heading to ICLR today and excited to be presenting two papers! 🇧🇷
If you’re interested in self-improving agents, lifelong learning, exploration, or just want to connect, I’d love to chat. DMs are open!
👇 I’ll be presenting:
📝 RLAD: Training LLMs to Discover Abstractions for Solving Reasoning Problems
📅 04/23, 3:15 PM | Pavilion 3, Poster #1413
📝 FSPO: Few-Shot Optimization of Synthetic Preferences Personalizes to Real Users
📅 04/23, 10:30 AM | Pavilion 3, Poster #1509
9
56
8,739
Excited to share our new work on anticipating scientific insights!
Given prior papers, can a model predict the core idea of the next paper?
With GIANTS-4B, we start to model the “conceptual leap” behind discovery.
Amazing work led by @JoyHeYueya @Anikait_Singh_
Scientists often make breakthroughs by synthesizing ideas across papers. In our new paper, we ask whether a language model can anticipate this process: given two parent papers, can it generate the core insight of a future paper built on them? 🧵⬇️
4
337
Ge Gao retweeted
Scientists often make breakthroughs by synthesizing ideas across papers. In our new paper, we ask whether a language model can anticipate this process: given two parent papers, can it generate the core insight of a future paper built on them? 🧵⬇️
18
91
723
188,774
Please stop by if you are interested! I’d love to chat!
And our 3rd workshop poster is on off policy RL selection!
ARLET RL Workshop (Rm 31abc) 3:30 p.m
Behavior-Aware Off-Policy Selection in High-Stake Human-Centric Environments
w/G.Gao, Aishwarya Mandyam, Joy He-Yueuya & Min Chi #NeurIPS2025
225
Ge Gao retweeted
And our 3rd workshop poster is on off policy RL selection!
ARLET RL Workshop (Rm 31abc) 3:30 p.m
Behavior-Aware Off-Policy Selection in High-Stake Human-Centric Environments
w/G.Gao, Aishwarya Mandyam, Joy He-Yueuya & Min Chi #NeurIPS2025
1
13
2,154
Ge Gao retweeted
I’m recruiting several PhD students at Carnegie Mellon University!
If you’re interested in LLM reasoning, agents, or diffusion language models, consider applying to the CMU ECE PhD program.
Applications are due Dec 15.
ece.cmu.edu/admissions/gradu…
10
104
464
63,718
Ge Gao retweeted
Could the way a student interacts with educational software in the first few hours predict their success months later? A new study suggests it's possible. hai.stanford.edu/news/assess…
5
11
1,924
Ge Gao retweeted
We’ve released updated 2024 lectures from my Stanford RL CS234 class! shorturl.at/pf427
This includes a guest lecture on Direct Preference Optimization (DPO) from 1st authors @rm_rafailov @archit_sharma97 @ericmitchellai tinyurl.com/3kr4czth Thanks to @StanfordOnline!
5
112
597
46,063
(1/n) How can we create personalized interventions in healthcare or education without prior data from each new participant? What if an AI could select the most suitable strategy at "first glance" to maximize outcomes for diverse individuals? #rl #Healthcare #Education
1
428
(3/n) Tested in intelligent tutoring and sepsis treatment, FPS can improve outcomes by up to 208% in education in a retrospective study and outperformed conventional methods in healthcare. FPS is a leap in adapting RL policies to real-world, personalized applications!
1
90
(4/n) An Arxiv version is available online: arxiv.org/html/2410.20017v1. More details to be released. Big shout out to the co-authors who devoted tremendous efforts toward making this real!! Xi Yang, @GaoQitong, Song Ju, @miroslav_pajic, and Min Chi!
77
Check out our #NeurIPS2023 paper about off-policy evaluation for human feedbacks!
#RLHF pioneers large-scale autonomous agent training with HF, e.g., for #LLMs. Curious if one could estimate the HF without the need of laboriously labeling it? Our work on off-policy evaluation for human feedback (OPEHF) accepted at #NeurIPS explains how! arxiv.org/pdf/2310.07123.pdf
337
Ge Gao retweeted
Thanks for sharing. I've added it as a reference in here:
yisongyue.medium.com/checkli…
4
5
RT @SIGCSE_TS: Morning, everyone! The COMPLETE #SIGCSE2021 program is now LIVE! Check it out now.
sigcse2021.sigcse.org/schedu… t.co/GdZ…
3