Vision-Predictor-Action: Toward Efficient Embodied AI
Overview
A new method, Vision-Predictor-Action (VPA), uses neuro-symbolic JEPA predictors to improve robot control with faster inference and lower parameter count, achieving 40% success on LIBERO-10 with a single GPU and under 20M parameters.
Generated from attributed reports · Updated 1 hours ago
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 · evaluation episodes: 200 other · Basis not reported
“mera reference configuration takes 10-14 ms on an Apple M4; and trained only on LIBERO-10 demonstrations in under three hours on a single GPU, a two-camera policy succeeded in 40.0% of 200 evaluation episodes. These preliminary results from a single training run suggest the approach is workable. Code and weights are available at https://github.com/ylee-sling/Vision-Predictor-Action and https://huggingface.co/ylee-sli”
Exact source · revision 1Source owner not reported
Reported quantity · training time: 3 hours · Basis not reported
“mera reference configuration takes 10-14 ms on an Apple M4; and trained only on LIBERO-10 demonstrations in under three hours on a single GPU, a two-camera policy succeeded in 40.0% of 200 evaluation episodes. These preliminary results from a single training run suggest the approach is workable. Code and weights are available at https://github.com/ylee-sling/Vision-Predictor-Action and https://huggingface.co/ylee-sli”
Exact source · revision 1Source owner not reported
Reported quantity · inference time per decision step: 10 other · Basis not reported
“mera reference configuration takes 10-14 ms on an Apple M4; and trained only on LIBERO-10 demonstrations in under three hours on a single GPU, a two-camera policy succeeded in 40.0% of 200 evaluation episodes. These preliminary results from a single training run suggest the approach is workable. Code and weights are available at https://github.com/ylee-sling/Vision-Predictor-Action and https://huggingface.co/ylee-sli”
Exact source · revision 1Source owner not reported
Reported quantity · checkpoint size: 75 other · Basis not reported
“horizon. We analyze inference complexity, regularization against representation collapse, and conditions under which latent sub-goal thresholding limits error accumulation across task phases. As an initial proof of concept, we evaluated VPA on the LIBERO-10 suite. The trainable model has under 20M parameters (checkpoint under 75 MB, excluding a frozen 63.4M-parameter CLIP text encoder); a decision step of a single-ca”
Exact source · revision 1Source owner not reported
Artifact availability · code: available · Differing source assertions
“mera reference configuration takes 10-14 ms on an Apple M4; and trained only on LIBERO-10 demonstrations in under three hours on a single GPU, a two-camera policy succeeded in 40.0% of 200 evaluation episodes. These preliminary results from a single training run suggest the approach is workable. Code and weights are available at https://github.com/ylee-sling/Vision-Predictor-Action and https://huggingface.co/ylee-sli”
Exact source · revision 1Source owner not reported
Artifact availability · code: available · Differing source assertions
“mera reference configuration takes 10-14 ms on an Apple M4; and trained only on LIBERO-10 demonstrations in under three hours on a single GPU, a two-camera policy succeeded in 40.0% of 200 evaluation episodes. These preliminary results from a single training run suggest the approach is workable. Code and weights are available at https://github.com/ylee-sling/Vision-Predictor-Action and https://huggingface.co/ylee-sli”
Exact source · revision 1Source owner not reported
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsVision-Predictor-Action: Towards Efficient Embodied AI via Neuro-Symbolic JEPA Predictors
The paper presents Vision-Predictor-Action (VPA), a method for efficient embodied AI that uses a neuro-symbolic JEPA predictor to generate robot actions. It replaces autoregressive decoders with a lightweight latent control loop, enabling faster inference and improved closed-loop control. Evaluated on LIBERO-10, the model achieves 40.0% success in 200 episodes with a single GPU and under 20M parameters.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).