Max Rudolph

rl + robotics, cs phd @UTAustin | AI/Robotics Research Scientist Intern @AIatMeta | prev @Amazon, @NASAJPL, bs ms @GeorgiaTech

[Max Rudolph @maxbrudolph]

We also release code for benchmarking!

https://t.co/u3WR9KNGja

[NYU Center for Data Science @NYUDataScience]

Simple deep RL was thought to fail in imperfect-info games like poker.

A new ICLR 2026 paper shows that with proper tuning, generic methods like PPO match or beat specialized approaches like fictitious play and counterfactual regret minimization.

https://t.co/RvkZ6XxR6J

1/2

[RL Beyond Rewards Workshop @RLBRew_RLC]

Two BIG updates for the RLBrew Workshop at #RLC2026! 📣

1️⃣ Dual submissions are welcome
2️⃣ We’ll be awarding a Best Paper RLBrew Award 🏆

You have 2 DAYS LEFT to submit — deadline: May 29!

Details: https://t.co/segLTne6Tp

[RL Beyond Rewards Workshop @RLBRew_RLC]

Reminder! RLBRew deadline in coming up in 7 days! Submit your works soon👩‍💻

Reminder that we accept under review papers! This is a good place to discuss your ideas and get feedback from the community

[RL Beyond Rewards Workshop @RLBRew_RLC]

We are extending the submission deadline to May 29, 2026 (AoE). Submit your ideas soon!

[Ziyan "Ray" Luo @RayZiyan41307]

We welcome diverse perspectives. Your contribution would be invaluable! 🙏

[Tiffany Ding @tifding]

Come say hi at my @iclr_conf poster (Friday 10:30am, Pavilion 3 #316) to chat about prediction sets or why @PlantNetProject is my favorite plant identification app!

[Max Rudolph @maxbrudolph]

Presenting Reevaluating Policy Gradient Methods in IIG @iclr_conf ! Come by poster 4418 tomorrow at 10:30 if you want...the (exp-a-) spiel

Work done with @nathanlichtl @___sobhan___ @alexandrebayen , @zicokolter, @yayitsamyzhang @gabrfarina @EugeneVinitsky and @ssokota

[Max Rudolph @maxbrudolph]

We ran thousands of sweeps to compare RL algos for imperfect information games and found preliminary evidence for the Policy Gradient Hypothesis:

With proper tuning, generic PG (PPO, etc.) methods are highly competitive in IIGS.

Check out the full paper: https://t.co/Cw4krtwVWe

[Ziyan "Ray" Luo @RayZiyan41307]

🔥 Thrilled to announce the Continual Reinforcement Learning (CRL) Workshop @continual_learn, at
@RL_Conference 2026 in Montreal, Canada!

📣 We welcome submissions on broad topics of continual RL. Interested in submitting or reviewing? Check out our website for more details!

[Max Rudolph @maxbrudolph]

@RL_Conference is one of my favorite conferences to submit, go, and review. I suspect it will be my favorite to host a workshop too!

Great speaker lineup, submit your work!

[RL Beyond Rewards Workshop @RLBRew_RLC]

Scalar, well-defined, easy-to-optimize rewards aren’t always available in real-world interaction data–yet that data is crucial for scaling general-purpose agents. Excited to announce the 3rd edition of RLBRew: Towards Scalable General-Purpose Agents at @RL_Conference 2026!

[Max Rudolph @maxbrudolph]

@harshit_sikchi The cafe scene NEEDS you back in Austin.

[Harshit Sikchi @harshit_sikchi]

Have seen the most thoughtful reviews so far in RLC. Watching the review quality go ⬇️ and conference size ⬆️at ICLR, Neurips past years they seem like the wrong venue to facilitate discussions on RL. Consider submitting!

[Max Rudolph @maxbrudolph]

Loved working with @akhil_bagaria et al.! Some amazing applications of RL to real world problems that we all want! Highly recommend!

[Akhil Bagaria @akhil_bagaria]

My RL team at @amazon NYC is looking for summer 2026 PhD interns! We apply RL to Amazon's supply chain and do publishable, open-ended research in meta learning, multi-agent RL, constrained RL, exploration, and LLMs+RL. Interested? Email your CV to [email protected] by Nov 30th!

[Max Rudolph @maxbrudolph]

Go see Harshit talk about this work! I think its one of the coolest projects i've been involved with during my PhD thus far.

[Harshit Sikchi @harshit_sikchi]

Announcing RLZero for Generalist Agents at #NeurIPS2025. To our knowledge, the first to enable all of:
💬 Language → behavior (zero-shot)
🎥 Video → behavior (zero-shot, cross-embodiment)
🧠 One Behavioral Foundation Model for many tasks

From instructions & demos to actions—no fine-tuning/test-time learning. 🚀 Below are examples what RLZero can do for humanoids:

[Max Rudolph @maxbrudolph]

Very cool work! The human video-to-robot behavior gap is a tough one to bridge!

[Rutav @rutavms]

Intelligent humanoids should have the ability to quickly adapt to new tasks by observing humans

Why is such adaptability important?
🌍 Real-world diversity is hard to fully capture in advance
🧠 Adaptability is central to natural intelligence

We present MimicDroid 👇

🌐 https://t.co/J8XpND9j1j

[Harshit Sikchi @harshit_sikchi]

RLZero will be presented at @NeurIPSConf 2025 . Learn more about the work in the thread below:

[Max Rudolph @maxbrudolph]

@akhil_bagaria At least the supply of PhD students who study RL is stable

[Max Rudolph @maxbrudolph]

@doomie @dwarkesh_sp @svlevine If latent, multi-step prediction through auto-regression or multi-step latent representation 👀

[Max Rudolph @maxbrudolph]

@danfei_xu to your point, I think what has made the Jet Propulsion Lab so good at space robotics is their heavy and meticulous focus on systems engineering.

[Max Rudolph @maxbrudolph]

@daphne_cor In addition to this, I like to jot down what to-dos I actually accomplished from the previous day too

[Max Rudolph @maxbrudolph]

@harshit_sikchi @RL_Conference Where's the fine-tuned cartwheel?