[Independent Variable Robotics Releases Cross-Embodiment Action Tokenizer X-Tokenizer] According to Golden Finance, on June 30, Independent Variable Robotics released the cross-embodiment action tokenizer X-Tokenizer. It is reported that X-Tokenizer adopts an 'encoder-semantic residual quantization (SRQ)-decoder' architecture, replacing the traditional standard residual vector quantization (RVQ) with SRQ. This layer separates action intentions and has been pre-trained on 2.4 million trajectories (including 2 billion action frames) covering 17 robotic arm series.