Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

Argus: an open-source robotics data annotation and quality pipeline. With new frontier VLMs like GPT-6 Astra, we can generate rich, high-quality annotations for robotics data for both training and dataset analysis. Argus delivers detailed, timestamp-level annotations while catching issues like mislabeled instructions, sped-up recordings, swapped camera streams, and unflagged...

89,243 Aufrufe • vor 1 Monat •via X (Twitter)

33 Kommentare

Profilbild von calix
calixvor 1 Monat

Argus takes in any episode in its raw format and outputs the following: - A densely annotated timeline with all robot or gripper actions, state changes, task progression percentages, and more - A record for every operator mistake, along with notes on the timing of the failure, the nature of the failure, the operator’s recovery response, and the ideal fix - Any metadata mistakes - A severity rating for each issue based on the predicted impact on training - Any subgoals and the completion outcome - The goal frame if it is achieved - Overall grade on completion along with a short rationale review

Profilbild von calix
calixvor 1 Monat

Argus can also be used on inference and tele-op trajectories captured in production. Here’s an example of a fully annotated trajectory from one of our production facilities.

Profilbild von calix
calixvor 1 Monat

Argus is a model-agnostic pipeline and can run on any sufficiently capable VLM. In our test comparisons, Astra and GPT-6.1 Sol produced the densest annotations, which we’ve found is highly correlated with overall accuracy.

Profilbild von calix
calixvor 1 Monat

Of the 3,546 episodes we labelled across nine datasets, 27% have at least one issue of medium severity or above, meaning part or all of the episode teaches the model something wrong. We’re releasing the full audit and more than 150K annotations at

Profilbild von calix
calixvor 1 Monat

Read the full analysis at and run Argus on your data at For collaborations or questions, contact us at [email protected]. Credit to @ericli_ for undertaking this project.

Profilbild von Henry Lien
Henry Lienvor 1 Monat

@PantheonInc Eric 🚀

Profilbild von koko 𝑥𝑠
koko 𝑥𝑠vor 1 Monat

@PantheonInc 🚀

Profilbild von andrew kim
andrew kimvor 1 Monat

@PantheonInc Incredible

Profilbild von Varun Nair
Varun Nairvor 1 Monat

@PantheonInc Pushing the frontier! VLMs are improving so fast, habit. Great harnesses and clarity for training ready data is key to extract signals from raw data and train the right policies!

Profilbild von Mohamed Zidan Cassim
Mohamed Zidan Cassimvor 1 Monat

@PantheonInc chef Li @ericli_

Profilbild von carlo agostinelli
carlo agostinellivor 1 Monat

@PantheonInc 🚢

Profilbild von Sambhav Gupta
Sambhav Guptavor 1 Monat

@PantheonInc Eric li rolling the robotics boulder up the hill

Profilbild von Xiao-ke (XK)
Xiao-ke (XK)vor 1 Monat

@PantheonInc @ericli_ 🚀🚀🚀

Profilbild von sean jeong
sean jeongvor 1 Monat

@PantheonInc 🔥

Profilbild von Rich
Richvor 1 Monat

@PantheonInc Mr Li

Profilbild von Ishan
Ishanvor 1 Monat

@PantheonInc So cool

Profilbild von Yang Fan Yun
Yang Fan Yunvor 1 Monat

@PantheonInc very very cool stuff @calixo888 @nettedjay 😎

Profilbild von Tian Fang
Tian Fangvor 1 Monat

@PantheonInc Very cool.

Profilbild von Nikhil Suresh
Nikhil Sureshvor 1 Monat

@PantheonInc 🚢🚢🚢

Profilbild von Felix Huang
Felix Huangvor 1 Monat

@PantheonInc Fire contributions for the community

Profilbild von Hudzah
Hudzahvor 1 Monat

@PantheonInc beautiful

Profilbild von ChrisPy
ChrisPyvor 1 Monat

@PantheonInc 🧑‍🍳

Profilbild von Sufyan Osamah
Sufyan Osamahvor 1 Monat

@PantheonInc Swapped camera streams and sped up recordings silently ruin a training run. Good that this is open source.

Profilbild von majjigi
majjigivor 1 Monat

@PantheonInc damn, this is amazing.

Profilbild von Lucas
Lucasvor 1 Monat

@PantheonInc Congrats Calix!

Profilbild von Arin
Arinvor 1 Monat

@PantheonInc 27% of episodes having training-affecting issues is a wild number for datasets people assumed were clean

Profilbild von joseph
josephvor 1 Monat

@PantheonInc cooking

Profilbild von Federico Martelli
Federico Martellivor 1 Monat

@PantheonInc catching swapped camera streams and mislabeled instructions before training saves more than another round of fine tuning, factory data has the same mess

Profilbild von Alexander
Alexandervor 1 Monat

@PantheonInc Super cool ! Have you guys looked at how much variation there is in SR of VLA style models when there is variance in the task description?

Profilbild von ^_^
^_^vor 1 Monat

@PantheonInc That's amazing!

Profilbild von Jack
Jackvor 1 Monat

Generated annotations need a sampling audit or you inherit the model's blind spots at full scale. The failure mode is not random noise, it is consistent error, which looks exactly like clean data until something downstream learns the same mistake. Worth holding back a human labelled slice purely to measure drift against.

Profilbild von Anna /✦
Anna /✦vor 1 Monat

@PantheonInc sharing with my friend in robotics!

Profilbild von Kifayath Ulla
Kifayath Ullavor 1 Monat

@PantheonInc WOAHH

Ähnliche Videos

🔥 JUST IN: Open-source robotics dataset from 100% real-world scenarios! 🤯 Chinese robotics company AGIBOT just released AGIBOT WORLD 2026, an open-source dataset systematically covering key embodied AI research directions. Built entirely from real-world environments: commercial spaces, and homes. Collected using AGIBOT G2 robots in free-form collection mode, providing structured, accurately annotated, high-quality data. Digital twin technology creates 1:1 scale replicas in simulation matching the real environments. Both real-world and simulation data are open-sourced. The AGIBOT G2 platform collects multiple data types simultaneously: RGB(D) cameras, tactile sensors, force sensors, LiDAR, IMU, and full-body joint states. Whole-body control coordinates arms, waist, and hands for complex tasks. First-person teleoperation lets operators control the robot from its perspective. The tasks covered are fine-grained manipulation, ultra-long-horizon tasks, spatial navigation, dual-arm coordination, and multi-agent/human-robot collaboration. The dataset includes error-recovery trajectories with annotations. Most datasets only show successful demonstrations. AGIBOT includes failures and how the robot recovers, teaching models how to handle mistakes. After collection, data is tested through policy training and real-robot deployment to ensure quality. Then processed through industrial quality control with multiple screening and cleaning rounds. Making it open-source accelerates embodied AI research by giving researchers access to high-quality real-world robot data at scale. 🇨🇳 Learn more here: ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

40,583 Aufrufe • vor 7 Monaten