RoboDrop: Curating VLA Post-Training Data via Local Gradient Compatibility
Overview
RoboDrop is a data curation framework for Vision-Language-Action (VLA) models that evaluates data quality by comparing gradient directions between candidate samples and clean validation data. It improves success rates on real robots from 35.0% to 67.5% by automatically excluding corrupted episodes. The method uses local gradient compatibility to identify and remove suboptimal demonstrations, achieving high performance in both simulated and real-world tasks.
From Robotics — Robot learning news discovery
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Reported quantity · Reported success rate: 35 percent · Basis not reported · Differing source assertions
“each training sample points in the same direction as the gradient of a clean validation sample in a similar task state. It evaluates data quality online during a one-epoch warm-up and automatically excludes bad episodes. On a real robot, the average success rate improved from 35.0% before curation to 67.5% with automatic budget selection. Title: RoboDrop: Curating VLA Post-Training Data via Local Gradient Compatibili”
Exact source · revision 2Source owner not reported
Reported quantity · Reported success rate: 67.5 percent · Basis not reported · Differing source assertions
“each training sample points in the same direction as the gradient of a clean validation sample in a similar task state. It evaluates data quality online during a one-epoch warm-up and automatically excludes bad episodes. On a real robot, the average success rate improved from 35.0% before curation to 67.5% with automatic budget selection. Title: RoboDrop: Curating VLA Post-Training Data via Local Gradient Compatibili”
Exact source · revision 2Source owner not reported
Reported quantity · Reported success rate: 84 percent · Basis not reported · Differing source assertions
“ts: Improved corruption detection and policy performance Figure 5: Examples of suboptimal trajectories detected by RoboDrop For temporal corruption in LIBERO, while the unfiltered average success rate was 84.0% and 83.4% with 10% random exclusion, RoboDrop achieved 91.6%. For action corruption, it improved from 91.3% and 92.0% to 95.1%. The temporal corruption detection AUROC for LIBERO-10 was 96.1%, surpassing the s”
Exact source · revision 2Source owner not reported
Reported quantity · Reported success rate: 83.4 percent · Basis not reported · Differing source assertions
“ts: Improved corruption detection and policy performance Figure 5: Examples of suboptimal trajectories detected by RoboDrop For temporal corruption in LIBERO, while the unfiltered average success rate was 84.0% and 83.4% with 10% random exclusion, RoboDrop achieved 91.6%. For action corruption, it improved from 91.3% and 92.0% to 95.1%. The temporal corruption detection AUROC for LIBERO-10 was 96.1%, surpassing the s”
Exact source · revision 2Source owner not reported
Artifact availability · paper: available
“ty URL: https://arxiv.org/abs/2609.10021 Affiliation: Tsinghua University, Striding AI Authors: Runze Xu, Yuanfan Xu, Cuijie Xu, Shuang Dai, Yining Li, Yu Wang, Jincheng Yu Figure 1: RoboDrop auditing VLA data using local gradient compatibility Background: Even successful episodes contain local errors Robot demonstrations may contain unnecessary movements or misalignments in observations and actions, even if the task”
Exact source · revision 2Source owner not reported
Evidence dependency changes (2)
- 2026-10-30T20:11:34.985Z · evidence updated · source revision 2. Evidence extraction was updated; current source attributions are shown above.
- 2026-10-30T20:11:24.913Z · evidence updated · source revision 2. Evidence extraction was updated; current source attributions are shown above.
Report timeline
Follow attributed reports and material updates.
- Robotics — Robot learning news discoverySignalRoboDrop: Curating VLA Post-Training Data via Local Gradient Compatibility
RoboDrop is a data curation framework for Vision-Language-Action (VLA) models that evaluates data quality by comparing gradient directions between candidate samples and clean validation data. It improves success rates on real robots from 35.0% to 67.5% by automatically excluding corrupted episodes. The method uses local gradient compatibility to identify and remove suboptimal demonstrations, achieving high performance in both simulated and real-world tasks.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).