Haoru Xue

@HaoruXue

PhD @berkeley_ai | prev. intern @nvidia GEAR, @CMU_Robotics, @LeCARLab | Robot Learning, Humanoids

San Francisco Bay Area

haoruxue.github.io

Joined December 2023

512 Following

3,047 Followers

Replies Reposts Media

Pinned
Haoru Xue
@HaoruXue
Dec 2, 2025
Reality of robotics: humanoid kung fu is solved before they can open doors with RGB. Here we are. Introducing the frontier of sim2real at NVIDIA GEAR. 100% sim data. RGB input only. Code name: 𝗗𝗼𝗼𝗿𝗠𝗮𝗻. We are opening the sim-to-real door. doorman-humanoid.github.io 🧵
00:00
13
Haoru Xue
@HaoruXue
Oct 22
Roboticist: “Oh no I’m not buying for kids” - “Well actually I am”
Haoru Xue
@HaoruXue
Oct 20
This is just the start of a comeback…
zeji yi
@ZejiYi
Oct 20
Common belief: Robot policies need continuous action heads; discrete actions are 𝐥𝐞𝐬𝐬 𝐧𝐚𝐭𝐢𝐯𝐞 𝐚𝐧𝐝 𝐥𝐞𝐬𝐬 𝐩𝐫𝐞𝐜𝐢𝐬𝐞. ❗We find policies w/ the simplest discrete representations 𝐦𝐚𝐭𝐜𝐡 𝐨𝐫 𝐛𝐞𝐚𝐭 continuous ones—in sim & real, in single-step &
00:00
1
Haoru Xue
@HaoruXue
Oct 19
Code as policy means code ⊂ policy, not code = policy Robotics should not stop at Astra writing code. E2E is always the objective. Spoiler alert 👀
Haoyang Weng
@ElijahGalahad
Oct 18
My blog on System 0 in the Astra era: As spatial grounding (S1/2) improves, we need dynamics-grounded controllers (S0) and expressive interfaces that let Astra steer physical interaction. Hope this resonates or sparks disagreement. I’d love your thoughts! haoyangweng.pages.dev/blog/2026-10-s…
00:00
3
Haoru Xue
@HaoruXue
Oct 14
CaP++++ lmao
Biwei Huang
@huang_biwei
Oct 12
I have spent more than a decade studying causality. For me, causality is about understanding the rules that govern how actions change the world, and reasoning about their consequences before acting. Scaling has made today's models remarkably capable at spotting patterns and
Aether AI (Causal Intelligence)
00:00
Aether AI (Causal Intelligence)