🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Check out what I have been working on the past few months! SMODICE is a simple and versatile offline IL algorithm that is compatible with learning from observations, mismatched experts, and even just examples! Detailed tweet coming soon 📅

Does this mean I’m a “Twit” now?

Machine Learning is really good at making point predictions --- but it sometimes makes mistakes. How should we think about which predictions we should trust? In other words, what is the right way to think about the uncertainty of particular predictions? A thread about new work 🧵

I worked with Meta last year on using multi-armed bandits to dynamically improve their content moderation. Excited that the short paper is finally up! arxiv.org/abs/2211.06516 Was fun working on this with @vas1089, Osbert Bastani, Deeksha, Parikshit, Caner & many others at Meta CDS

The exploration-exploitation tradeoff in #RL raises concerns about exploration -- who does it impact and how much? In a recent #AISTATS paper, we show how to effectively "spread out" exploration across episodes (individuals) w/ only a small cost to regret: arxiv.org/abs/2110.13060

Excited to share our #ICML2023 paper ✨LIV✨! Extending VIP, LIV is at once a pre-training, fine-tuning, and (zero-shot!) multi-modal reward method for (real-world!) language-conditioned robotic control. Project: penn-pal-lab.github.io/LIV Code & Model: github.com/penn-pal-lab/L… 🧵:

Excited to share our work on using generative AI tutors in the classroom - we ran a RCT on 1000+ students at a large high school in collaboration with math teachers. We find that students use GenAI as a "crutch," harming their learning: papers.ssrn.com/sol3/papers.cf…

We ran a randomized controlled trial on 1000+ students at a large high school and found that GenAI can harm their learning if not used carefully, details below!

Our paper "Generative AI Can Harm Learning" was cited in the 2025 Economic Report of the President -- first paper I've written that someone outside of academia has read! Or at least the abstract 😂

Conformal prediction sets are a useful way to capture uncertainty for LLMs & deep learning models. But they're data-hungry! We propose a semi-bandit algo to learn these sets online. Check out our ICML Conference paper: arxiv.org/abs/2405.13268 Work led by Haosen Ge, Ph.D. & with Osbert Bastani