Linh Le @linhlpv

I was happy to give a more technical talk on how we might create an AI at RLC-2025 and AGI-2025 (video below). The Oak Architecture: A Vision of Super-Intelligence from Experience As AI has become a huge industry, to an extent it has lost its way. What is needed to get us back on…

Finding an ML summer school has never been easier Here is a GitHub repo with a comprehensive list, with 50+ ML summer (and winter) schools all over the world (link in comments) Some of them are even free, few even offer scholarship so you don't have to pay absolutely anything

🚀 I'm excited to share our new paper: SegDAC: Segmentation-Driven Actor-Critic for Visual Reinforcement Learning 🧠 SegDAC combines large vision models with online RL to reason about its environment at the object and sub-object level, avoiding noisy pixel-level reasoning. 🛠️…

Humanoids finally move like humans… and can do more than copy. [Details + demos in thread 👇] A new framework, BeyondMimic, shows how to learn naturalistic whole-body control from human motion. But then goes further by composing those skills into versatile, zero-shot…

At what point does perf optimization get ridiculous. During my PhD, everything was 500-5000 sps. Then I got 10k and was very proud. Then 100k in early versions of PufferLib. Then 1M in 2.0... and now we're at up to 6M productive SPS on some RL envs

Fine-tuning pre-trained robotic models with online RL requires a way to train RL with expressive policies Can we design an effective method for this? We propose EXPO, a sample-efficient online RL algorithm that enables stable fine-tuning of expressive policy classes (1/6)

✨Introducing SENSEI✨ We bring semantically meaningful exploration to model-based RL using VLMs. With intrinsic rewards for novel yet useful behaviors, SENSEI showcases strong exploration in MiniHack, Pokémon Red & Robodesk. Accepted at ICML 2025🎉 Joint work with @cgumbsch 🧵

missing ICML, and I used this week to write my first technical blog on some recent thoughts on two different roles of simulators in RL and the confusions/misconceptions around them. Comments welcome! nanjiang.cs.illinois.edu/2025/07/16/sim…

The paper: arxiv.org/abs/2502.03349

Everyone knows action chunking is great for imitation learning. It turns out that we can extend its success to RL to better leverage prior data for improved exploration and online sample efficiency! colinqiyangli.github.io/qc/ The recipe to achieve this is incredibly simple. 🧵 1/N

Warm-start RL (WSRL) can learn to control a real robot in under 20 minutes! Deep RL is getting really fast. Warm-start from offline data + super-efficient online learning is increasingly making real world RL not just practical but pretty easy.

Paul Zhou @zhiyuan_zhou_

2 months ago

Warm-start RL (WSRL) can learn to control a real robot in under 20 minutes! Deep RL is getting really fast. Warm-start from offline data + super-efficient online learning is increasingly making real world RL not just practical but pretty easy.

9 38 253 82K 128

Download Video

You don't _need_ a PhD (or any qualification) to do almost anything. A PhD is a rare opportunity to grow as an independent thinker in an academic environment, rather than immediately becoming a gear in a corporate agenda. It's definitely not for everyone!

Noam Brown @polynoamial

2 months ago

You don't _need_ a PhD (or any qualification) to do almost anything. A PhD is a rare opportunity to grow as an independent thinker in an academic environment, rather than immediately becoming a gear in a corporate agenda. It's definitely not for everyone!

199 208 4K 1.3M 897

📢📢 "Align Your Flow: Scaling Continuous-Time Flow Map Distillation" New flow map framework for state-of-the-art few-step generation, w/ the amazing @amsabour and @FidlerSanja. 🔥 Project page: research.nvidia.com/labs/toronto-a… 📜 Paper: arxiv.org/abs/2506.14603 🧵Thread below... (1/n)

🚨 Excited to share our new work: "Stable Gradients for Stable Learning at Scale in Deep Reinforcement Learning"! 📈 We propose gradient interventions that enable stable, scalable learning, achieving significant performance gains across agents and environments! Details below 👇

Self-supervised representation learning looks a bit like RL. What if we literally use RL as a SSL method for visual representations? Turns out that it works quite well. In new work by @its_dibya, we show how this can be done: dibyaghosh.com/annotation_boo…

Normally, changing robot policy behavior means changing its weights or relying on a goal-conditioned policy. What if there was another way? Check out DynaGuide, a novel policy steering approach that works on any pretrained diffusion policy. dynaguide.github.io 🧵

Our view on test-time scaling has been to train models to discover algos that enable them to solve harder problems. @setlur_amrith & @matthewyryang's new work e3 shows how RL done with this view produces best <2B LLM on math that extrapolates beyond training budget. 🧵⬇️…

1/ How should RL agents prepare to solve new tasks? While prior methods often learn a model that predicts the immediate next observation, we build a model that predicts many steps into the future, conditioning on different user intentions: chongyi-zheng.github.io/infom.

Hierarchical methods for offline goal-conditioned RL (GCRL) can scale to very distant goals that stymie flat (non-hierarchical) policies — but are they really necessary? Paper: arxiv.org/abs/2505.14975 Project page: johnlyzhou.github.io/saw/ Code: github.com/johnlyzhou/saw Thread ↓

Phaidra is hiring a Research Scientist to work on sequential decision-making problems. I'm at the RLDM conference in Dublin this week. If you're attending and would like to learn more about the role or the company, feel free to reach out! job-boards.greenhouse.io/phaidra/jobs/4…