Conceptvisual token pruning
Visual Token Pruning
A VLA inference acceleration technique that reduces the computational overhead of perception and inference latency by compressing the number of visual input tokens.
Timeline (last 90 days)
Oct 18
Existing VLA accelerators (such as action chunking, visual token pruning, and quantized compilation) generally report only average task success rates, masking the risk of cascading failures that can occur during closed-loop execution.
Unverified50%