Jie Zhang

AI agents can now generate and verify complex 3D worlds using executable code, outperforming leading models in human preference studies.

World action models do not need to render clean future video to generalize: conditioning on a single forward pass of pure noise captures virtually all the robustness benefits of multi-step diffusion at a fraction of the compute.

Standard on-policy RL curbs vision-language hallucinations through an uninformative brevity shortcut, a failure mode resolved here by pairing exhaustive concept verification with subsentence-level credit assignment.

Direct radar velocity measurements can anchor visual object permanence through complete optical dropouts and clutter, driving a massive 30-point MOTA leap in multimodal moving-object tracking.

Language-model judges may be fundamentally unreliable, with repeatability scores failing to meet established benchmarks, jeopardizing the integrity of AI evaluations.

Aggregating insights from neighboring users' reviews can transform how we tackle data sparsity in recommender systems, leading to better predictions and richer explanations.

Intent-targeted tools reveal that LLMs often signal harmful intentions before executing risky actions, enabling real-time intervention strategies.

IQACO not only boosts observation efficiency for agile satellites but also adapts dynamically to the unpredictable nature of maritime targets.