Shengju Qian

Surpassing existing methods, Orchestra-o1 achieves a 10.3% accuracy improvement on the OmniGAIA benchmark by enabling seamless collaboration across multiple modalities.

Standard RL rollouts can effectively provide world modeling supervision, leading to significant performance gains in language agents.

Distilling black-box video generators into autoregressive models doesn't require teacher scores or complex alignment—just cleverly paired rollouts and a discriminator.

Autoregressive Transformers can now generate modular 3D assets from text, opening new avenues for UGC and professional 3D content creation.