Loading video...

Video Failed to Load

Go Home

Introducing Axis Dataset V1 - the simulation dataset for scalable robot manipulation. Can noisy, crowdsourced simulation data support embodied pretraining? Yes—with enough coverage and diversity. Continual pretraining on AXIS dataset V1 lifts π0.5 from 83.9% to 88.8% on LIBERO-Plus, with performance improving consistently as the pretraining data scales from...

114,840 views • 3 months ago •via X (Twitter)

34 Comments

Stone Tao's profile picture
Stone Tao3 months ago

don’t want to put too much shade on the progress (i think sim datasets can be good ideas if executed well!), but improving on libero (which is also a sim benchmark) means very little unfortunately

Thor's profile picture
Thor3 months ago

Let's goo 🔥

Hazzy's profile picture
Hazzy3 months ago

you edit too much intern

KABILA's profile picture
KABILA2 months ago

Cookingg

Jontu's profile picture
Jontu2 months ago

This is a strong proof that diverse simulation data can meaningfully improve embodied AI.

boyBrands_'s profile picture
boyBrands_3 months ago

He has arrived🔥

Sami's profile picture
Sami3 months ago

So much excited for this

Apurbo's profile picture
Apurbo2 months ago

We are so excited about this 🤘

AL Shifat's profile picture
AL Shifat3 months ago

Axis going to make history

dviet.eth's profile picture
dviet.eth3 months ago

This is exactly the kind of infrastructure Physical AI needs. Quality data compounds, and so does robot intelligence. Excited to see Axis Dataset V1 keep growing 😆

Noodle's profile picture
Noodle3 months ago

cool to see those teleop sessions actually pushing the libero-plus numbers up

Acee (✱,✱)'s profile picture
Acee (✱,✱)3 months ago

Keep cooking 👌

Piggie 🧡|Bird's profile picture
Piggie 🧡|Bird3 months ago

Overall success rate increased by 5.8%.

0xmowra.eth's profile picture
0xmowra.eth3 months ago

Axis V1 dataset boosting robot learning with crowdsourced simulation data impressive

Red's profile picture
Red2 months ago

Great work

| MinhDucc | Shark's profile picture
| MinhDucc | Shark3 months ago

Axis LFG 😍

sasafonda✱,✱)'s profile picture
sasafonda✱,✱)3 months ago

gaxis

namnew's profile picture
namnew3 months ago

gAxis

VINAY (✱,✱)'s profile picture
VINAY (✱,✱)3 months ago

Axis cooking 🔥

savagezz's profile picture
savagezz3 months ago

cooking something

Hoangdong (✱,✱)'s profile picture
Hoangdong (✱,✱)3 months ago

gAxis

clover's profile picture
clover3 months ago

move gaxis

vinhle (✱,✱) 🍚 ⛓ ./'s profile picture
vinhle (✱,✱) 🍚 ⛓ ./3 months ago

Exciting to see the potential of crowdsourced simulation data in robotics, and impressive results from continual pretraining on Axis Dataset V1.

Aurora Lux's profile picture
Aurora Lux3 months ago

AXIS Dataset V1 is a strong step forward for embodied AI—scalable, diverse simulation data is proving to be a powerful foundation for better robot learning. 🤖🚀

cbeta.base.eth's profile picture
cbeta.base.eth3 months ago

lets goo team fight fight fight

rara's profile picture
rara3 months ago

gaxis lfg

Linh (✱,✱)'s profile picture
Linh (✱,✱)3 months ago

Axis Dataset V1 is impressive

Nova's profile picture
Nova3 months ago

The bot is terrible. Please update it. Its performance is like a primitive human living in a cave 😤

MAS_Enpici (✱,✱)'s profile picture
MAS_Enpici (✱,✱)3 months ago

big update and new insight

yellowopus's profile picture
yellowopus3 months ago

lets cooking bro

KADAFER's profile picture
KADAFER3 months ago

let him cook 🔥

dondonat (✱,✱)'s profile picture
dondonat (✱,✱)3 months ago

gaxis keep building

sagar's profile picture
sagar3 months ago

Let's goo 🙌

Gujil Ruipa's profile picture
Gujil Ruipa3 months ago

Open datasets like this can accelerate robotics research through collaborative innovation

Related Videos

In my past research experience, finding or developing an appropriate simulation environment, dataset, and benchmark has always been a challenge. Missing features, limited support, or unexpected bugs often occupied my days and nights. Moreover, current simulation platforms are relatively fragmented—making it challenging to replicate the success of the RT-X dataset in unifying community efforts. Introducing RoboVerse, we provide a unified platform, dataset, and benchmark for scalable and generalizable robot learning. We hope to build a shared foundation to combine the community efforts. RoboVerse includes: MetaSim: We carefully designed a configuration system and a universal interface to align current robotic simulators. With MetaSim, you can use any simulator with the same code—bringing together the community’s diverse efforts under one framework! RoboVerse Dataset and Benchmark: We unify popular simulation environments and benchmarks into a single cohesive system and introduce the RoboVerse dataset—a large-scale, high-quality synthetic dataset. Additionally, we propose a standardized benchmark across both imitation learning and reinforcement learning. A cool feature enabled by our unified framework: Hybrid Simulation! You can now integrate physics engines and renderers from different simulators—e.g., using MuJoCo precise physics with Isaac photorealistic rendering. This not only elevates simulation fidelity but also significantly enhances real-world transfer performance across complex robotic applications. Hopefully, our team’s efforts could serve the robotic community to thrive vibrantly in the years to come. RoboVerse is open-sourced🥳!!! Project Page: Documentation: Github Repo: Paper:

Haoran Geng

84,425 views • 1 year ago