Policy learning via reward optimization, from model-free to model-based methods.
Hexian Ni, Tao Lu, Yinghao Cai
CoRe combines code-based rewards with learned visual preferences so that neither reward source has to capture the entire task alone.
Christos Ziakas, Alessandra Russo, Avishek Joey Bose
Generative policies based on expressive model classes, such as diffusion-models and flow matching, are well-suited to complex control problems with highly multimodal action distributions.
Kinam Kim, Namiko Saito, Heecheol Kim et al.
Vision-Language-Action (VLA) models can generalize across diverse manipulation tasks, but their imitation-learning-based policies remain brittle in precise physical interactions due to compounding execution errors; Can a reinforcement learning policy trained purely in simulation improve the robustness of real-world VLAs zero-shot?
Kinam Kim, Namiko Saito, Heecheol Kim et al.
Vision-Language-Action (VLA) models can generalize across diverse manipulation tasks, but their imitation-learning-based policies remain brittle in precise physical interactions due to compounding execution errors; Can a reinforcement learning policy trained purely in simulation improve the robustness of real-world VLAs zero-shot?
Jonghoon Lee, Seong Hyeon Park, Byungwoo Jeon et al.
Vision-language-action (VLA) policies have shown strong potential for general-purpose manipulation, yet they often fail on novel, out-of-distribution objects whose appearance or geometry deviates from the training distribution.
Jonghoon Lee, Seong Hyeon Park, Byungwoo Jeon et al.
Vision-language-action (VLA) policies have shown strong potential for general-purpose manipulation, yet they often fail on novel, out-of-distribution objects whose appearance or geometry deviates from the training distribution.
Chengyue Huang, Khang Vo Huynh, Sebastian Elbaum et al.
Robotic manipulation is typically evaluated by task success, but successful completion does not guarantee safe execution. Many rl failures are temporal: a robot may touch a clean surface after contamination or release an object before it is fully inside an enclosure.
Hoang-Dung Bui, Abhish Khanal, Raihan Islam Arnob et al.
SAP pairs aerial scouts with ground robots to gather environmental information ahead of time, reducing mission time by 40% through information-theoretic POMDP planning.
Ismail Geles, Leonard Bauersfeld, Markus Wulfmeier et al.
The first demonstration of superhuman, safe, multi-agent drone racing using a MARL framework trained in simulation and transferred to real Crazyflie nano-drones.
Loizos Hadjiloizou, Rodrigo Pérez-Dattari, Noémie Jaquier
Formalizes cross-space symmetry compositions in robotics and derives equivariant network architectures that exploit multiple simultaneous symmetries for improved generalization.
Matthew M. Hong, Jesse Zhang, Anusha Nagabandi et al.
Fine-tuning pre-trained robot policies with reinforcement learning (RL) often inherits the bottlenecks introduced by pre-training with behavioral cloning (BC), which produces narrow action distributions that lack the coverage necessary for downstream exploration.
Zachary Kousik, Shanpu Fang, Rohan Thakker et al.
An open-source 3D-printed humanoid robot platform designed for reinforcement-learning based locomotion research with affordable hardware.
Yecheng Jason Ma, William Liang, Guanzhi Wang et al.
Eureka uses GPT-4 to write reward functions for RL environments, achieving human-level reward design on 29 tasks and enabling zero-shot sim-to-real transfer on Shadow Hand dexterous manipulation.
Wang J., Xiao B., Wang J., Heider P., Li H., Zhang Y.
An Isaac-Gym-based training framework for humanoid locomotion that enables zero-shot sim-to-real transfer on XBot-L/XBot-S robots via domain randomization.
Remi Cadene, Simon Alibert, Alexander Soare et al.
Hugging Face's LeRobot is an open-source PyTorch framework providing pretrained models, datasets, and training scripts for imitation and reinforcement learning on real robots, lowering the entry barrier to robot learning.
Nikita Rudin, Fabian Jenelten, Timon Homberger et al.
Unitree Go2 uses cross-modal vision + proprioception transformers trained in Isaac Gym to achieve agile navigation over rough terrain with minimal sim-to-real gap.
Elia Kaufmann, Mathias Gehrig, Philipp Foehn et al.
An end-to-end neural drone controller trained entirely in simulation flies acrobatic maneuvers in the real world with zero real-world fine-tuning, enabled by domain randomization and reinforcement learning.
Farama Foundation, Jordan Terry, Mark Towers et al.
Gymnasium is the maintained successor to OpenAI Gym, providing a standardized API for RL environments with 100+ built-in tasks, vectorized parallel execution, and native support for physics engines like MuJoCo, PyBullet, and IsaacGym.
Jiayuan Gu, Sean Xiang, Stone Tao et al.
ManiSkill is a GPU-parallelized robotics simulation benchmark with 20+ manipulation tasks, PartNet-Mobility assets, and unified observation/action spaces, supporting RL, IL, and VLA training at millions of steps per hour.
Nikita Rudin, David Hoeller, Philipp Reist et al.
Legged Gym is an open-source training framework built on NVIDIA Isaac Gym that enables massively parallel deep RL for legged robot locomotion, achieving policy training times of minutes rather than days.
Nikita Rudin, David Hoeller, Marco Hutter
RSL-RL is a lightweight, high-performance PyTorch library implementing PPO tailored for locomotion. Used as the default RL backend in Legged Gym, it enables training thousands of agents in parallel on GPU.
Bokui Shen, Fei Xia, Chengshu Li et al.
iGibson is a photorealistic, physics-based indoor simulation platform with 15 interactive scenes, 500+ object categories, and GPU-accelerated rendering, enabling RL and VLA training for mobile manipulation and embodied AI.
Viktor Makoviychuk, Lukasz Wawrzyniak, Yunrong Guo et al.
A GPU-accelerated physics simulation framework that enables thousands of RL environments to run in parallel on a single GPU, eliminating CPU-GPU transfer bottlenecks.
Read these papers in order to build expertise in Reinforcement Learning.
Hexian Ni, Tao Lu et al. · 2026
Christos Ziakas, Alessandra Russo et al. · 2026
Kinam Kim, Namiko Saito et al. · 2026
Kinam Kim, Namiko Saito et al. · 2026
Jonghoon Lee, Seong Hyeon Park et al. · 2026
Jonghoon Lee, Seong Hyeon Park et al. · 2026
…and 17 more papers.
