Our work on generalized legged locomotion published in Science Robotics

www.science.org/doi/10.1126/...

It might be a bit late but here's our work in L4DC 2025 on bridging agility, safety, and adaptivity.

Compared to prior work (ABS) that safeguards agile locomotion, this one further explicitly estimates the change in dynamics and does adaptive shielding.

adaptive-safe-locomotion.github.io

Anyone else feel that academic people are not very active here, and even the following feed are always filled with anti-Musk/Trump political posts?

discord.gg/Xfxau8me

we just initiated a discord server on robot locomotion because we don't know any. Welcome to join!

Starting my PhD soon at ETHz... on I don't know what will happen but it's always on autonomy and legged robots.

The key to achieving horizon generalization is *planning invariance*. A policy is planning invariant if decomposing tasks into simpler subtasks doesn't improve performance. We prove planning invariance can enable horizon generalization. 3/

Very happy to show that we can do RL finetuning on 405B models with open-source code, beat Llama 405B instruct with their base model, and beat DeepSeek V3 too. Enjoy building off this teams hard work.

Here's Tulu 3 405B. A holiday present from @hamishivi.bsky.social @vwxyzjn.bsky.social and team.

Collection of biped locomotion+ RL papers

github.com/zita-ch/bipe...

Will (try to) begin work in 2025 (let's don't talk about exact dates).

🍏 New preprint alert! 🍏
PoM: Efficient Image and Video Generation with the Polynomial Mixer
arxiv.org/abs/2411.12663
This is my latest "summer project" and it was so big I had to call in reinforcements (Thanks @nicolasdufour.bsky.social)

TL;DR Transformers are for boomers, welcome to the future
🧵👇

Self intro with videos

I'm a Ms student (2022fall-2025spring) in #robotics at ETH Zurich.

I work on control and navigation for legged robots (quadruped, humanoid, wheeled-legged).

My focus is learning for performant and reliable legged systems.

My website: zita-ch.github.io