Embodiment-Classifier Guidance: Transferring Frozen VLAs Across Robot Dynamics
Overview
This research proposes a method to adapt frozen vision-language agents (VLAs) to new robot embodiments without additional demonstrations. By training a lightweight embodiment classifier to distinguish successful action chunks on source and target bodies, the method nudges the denoising steps of the original policy with the classifier's gradient, improving success rates on transfer tasks. The approach shows promise in simulation but has limitations when data is insufficient or motions are infeasible.
From arXiv Robotics — research abstracts
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Environment: simulation
“aises $\pi_0$'s average success on the target bodies from 72% to 80% across 12 suite-controller pairs, with a 31% relative reduction in failures (11 of 12 gain; 5 significant after Holm correction). In simulation, guidance is data-efficient, needing few target-body rollouts per task, which makes it promising for sim-to-real transfer, where rollouts are costly. Guidance is mixed on weak wrists and harmful on a locked”
Exact source · revision 1Source owner not reported
Reported quantity · number of tasks: 12 other · Basis not reported
“aises $\pi_0$'s average success on the target bodies from 72% to 80% across 12 suite-controller pairs, with a 31% relative reduction in failures (11 of 12 gain; 5 significant after Holm correction). In simulation, guidance is data-efficient, needing few target-body rollouts per task, which makes it promising for sim-to-real transfer, where rollouts are costly. Guidance is mixed on weak wrists and harmful on a locked”
Exact source · revision 1Source owner not reported
Reported quantity · number of successes: 11 other · Basis not reported
“aises $\pi_0$'s average success on the target bodies from 72% to 80% across 12 suite-controller pairs, with a 31% relative reduction in failures (11 of 12 gain; 5 significant after Holm correction). In simulation, guidance is data-efficient, needing few target-body rollouts per task, which makes it promising for sim-to-real transfer, where rollouts are costly. Guidance is mixed on weak wrists and harmful on a locked”
Exact source · revision 1Source owner not reported
Reported quantity · number of significant results: 5 other · Basis not reported
“aises $\pi_0$'s average success on the target bodies from 72% to 80% across 12 suite-controller pairs, with a 31% relative reduction in failures (11 of 12 gain; 5 significant after Holm correction). In simulation, guidance is data-efficient, needing few target-body rollouts per task, which makes it promising for sim-to-real transfer, where rollouts are costly. Guidance is mixed on weak wrists and harmful on a locked”
Exact source · revision 1Source owner not reported
Evidence dependency changes (2)
- 2026-10-30T22:34:46.464Z · evidence updated · source revision 1. Evidence extraction was updated; current source attributions are shown above.
- 2026-10-30T22:34:30.709Z · evidence updated · source revision 1. Evidence extraction was updated; current source attributions are shown above.
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsSignalEmbodiment-Classifier Guidance: Transferring Frozen VLAs Across Robot Dynamics without Demonstrations
This research proposes a method to adapt frozen vision-language agents (VLAs) to new robot embodiments without additional demonstrations. By training a lightweight embodiment classifier to distinguish successful action chunks on source and target bodies, the method nudges the denoising steps of the original policy with the classifier's gradient, improving success rates on transfer tasks. The approach shows promise in simulation but has limitations when data is insufficient or motions are infeasible.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).