Robotic Systems Lab

@leggedrobotics · Zurich, Switzerland · joined 07 Mar 2016

The Robotic Systems Lab designs machines, creates actuation principles, and builds up control technologies for autonomous operation in challenging environments.

38 268Followers

171Following

704Posts total

130KViews on collected posts

โพสต์ล่าสุด

Robotic Systems Lab@leggedrobotics22 Sep

VLMs don't understand a robot's body. How can a robot learn from its failures without retraining? PragmaBot: online in-context learning from real-world experience. The robot reflects on failures, stores lessons in memory, and retrieves them for new tasks. RA-L · #IROS2026 https

0:42

4.2K views · 51 likes · 7 reposts · 1 replies Open on X →

Robotic Systems Lab@leggedrobotics15 Sep

We are releasing EgoHTR, a dataset with both human motions and terrain references, accepted @corl_conf. 📖 Paper: https://t.co/LHEQpqFrXG 🌐 Project Page: https://t.co/x12lTzVTDf • 55 scene-aligned sequences • 150k+ frames • rough-terrain environments • multi-modal 3D scene https

2:05

11.1K views · 42 likes · 9 reposts · 3 replies Open on X →

Robotic Systems Lab@leggedrobotics10 Jun

@leggedrobotics Thanks for this! I want to start looking into off policy methods for humanoids and this is very useful 🙏🏼 1.4K views · 4 likes · 1 reposts · 1 replies Open on X →

Robotic Systems Lab@leggedrobotics10 Jun

This project is led by Gianluca Sabatini, supported by Chenhao Li @breadli428 and Marco Hutter @leggedrobotics. We thank Clemens Schwarke's implementation insights. Check out the paper! https://t.co/aVPFfzRecM 1.2K views · 8 likes · 0 reposts · 0 replies Open on X →

Robotic Systems Lab@leggedrobotics10 Jun

The implementation extends RSL-RL with same APIs, same training pipeline, same wrappers. ✅Swap the agent config, and you're running RSL-RL-SAC. Multi-GPU, RND, and symmetry augmentation are included out of the box. https://t.co/muvEjmdFYl 724 views · 7 likes · 0 reposts · 1 replies Open on X →

Robotic Systems Lab@leggedrobotics10 Jun

This project focuses on understanding the performance gap between PPO and SAC in massively parallel robot learning. While efforts like FlashSAC @hojoon_ai and FastSAC @younggyoseo explore separate development, RSL-RL-SAC is made to stay close to the widely used RSL-RL codebase.

1.6K views · 5 likes · 0 reposts · 1 replies Open on X →

Robotic Systems Lab@leggedrobotics10 Jun

On humanoid tasks, SAC actually outperforms PPO. Entropy-driven exploration pays off in dense reward settings. Unitree G1: https://t.co/wLgAKplhqb

780 views · 4 likes · 0 reposts · 1 replies Open on X →

Robotic Systems Lab@leggedrobotics10 Jun

By adapting SAC for massively parallel simulation in IsaacLab, we match PPO across quadruped locomotion tasks — with a single set of hyperparameters, NO reward retuning needed. Unitree Go2: https://t.co/DYbufekzyg

853 views · 3 likes · 0 reposts · 1 replies Open on X →

Robotic Systems Lab@leggedrobotics10 Jun

To make SAC work well, we crystallize four major factors that matter: ✅Right-sizing the action space ✅Treating timeouts as timeouts, not failures ✅Smoother targets via n-step returns ✅Starting exploration where it should 1.4K views · 9 likes · 0 reposts · 3 replies Open on X →

Robotic Systems Lab@leggedrobotics10 Jun

PPO has been the go-to algorithm for training robots in simulation. SAC is more sample-efficient in theory, but consistently fell short in practice. And while PPO thrives where data is cheap, it hits a hard wall when moving to real-robot learning. 🔥We set out to close that gap.

13.2K views · 23 likes · 3 reposts · 1 replies Open on X →

Robotic Systems Lab@leggedrobotics10 Jun

PPO has long dominated robot locomotion training in simulation. SAC, despite its sample efficiency, couldn't keep up. We analyze why: 🔗https://t.co/w8cR5lgxjf 🔥Integrated into RSL-RL, our approach requires only minimal changes, making SAC a drop-in alternative out of the box. h

0:19

45.6K views · 343 likes · 42 reposts · 7 replies Open on X →

Robotic Systems Lab@leggedrobotics1 Jun

We’re excited to be receiving one of these platforms for our research. Looking forward to exploring what we can build with it and contributing to the next wave of humanoid robotics. 7.6K views · 50 likes · 4 reposts · 3 replies Open on X →

Robotic Systems Lab@leggedrobotics19 May

Advancing dexterous manipulation through scalable visual sim-to-real transfer. We are excited to share our RSS paper, “ViserDex: Visual Sim-to-Real for Robust Dexterous In-hand Reorientation.” 🌐 Project page: https://t.co/eFi7QtKee9 1/N 🧵 https://t.co/YbKY5h6wmG

0:30

40.4K views · 199 likes · 20 reposts · 6 replies Open on X →

เทียบกับบัญชีขนาดเดียวกัน

2 โพสต์จาก 90 วันที่ผ่านมา เทียบกับช่วง 10K–100K ผู้ติดตาม แสดงในวงกว้าง แต่มีผู้ชมตอบสนองน้อย.

ยอดดูมัธยฐาน7 649บัญชีนี้992ค่ามัธยฐานของ 10K–100K

การเข้าถึง, %19.99%บัญชีนี้3.89%ค่ามัธยฐานของ 10K–100K

การมีส่วนร่วม, %0.97%บัญชีนี้1.54%ค่ามัธยฐานของ 10K–100K

ตัวชี้วัด	บัญชีนี้	ค่ามัธยฐานของ 10K–100K	อัตราส่วน
ยอดดูมัธยฐานต่อโพสต์	7 649	992	7.71×
การเข้าถึง (ยอดดู ÷ ผู้ติดตาม)	19.99%	3.89%	5.14×
อัตราการมีส่วนร่วม	0.97%	1.54%	0.63×

บัญชีอื่นในช่วงนี้ → เปรียบเทียบกับบัญชีอื่น → ค่าอ้างอิงเหล่านี้คำนวณอย่างไร →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

40.4K19 May

7.6K1 Jun

45.6K10 Jun

13.2K

1.4K

853

780

1.6K

724

1.2K

1.4K

11.1K15 Sep

4.2K22 Sep

Last 13 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.57%19 May

0.76%1 Jun

0.87%10 Jun

0.21%

0.86%

0.47%

0.64%

0.38%

1.10%

0.67%

0.44%

0.50%15 Sep

1.43%22 Sep

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.54%.

What the audience does

Likes55.4%748 in total

Reposts6.4%86 in total

Replies2.1%29 in total

Quotes1.2%16 in total

Bookmarks34.9%472 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

โพสต์ล่าสุด

เทียบกับบัญชีขนาดเดียวกัน

Growth & engagement

Views per post

Engagement rate per post

What the audience does

บัญชีที่คล้ายกัน