Jun Xiao

A training-free framework that reformulates failure recovery as test-time scaling over causal histories, which decomposes recovery into three coupled decisions: when to revise the causal history, where to recover a reliable history prefix, and which history configuration best supports subsequent execution.

This work proposes EvoSkill-GUI, a training-free framework in which each skill is a structured multi-file package containing retrieval metadata, executable plans, backup localization, failure-recovery rules, accessibility utilities, and failure cases.

TTPO achieves label-supervised performance without any ground-truth labels, outperforming traditional methods on key benchmarks.

Event Tokens can dramatically enhance LLM recommendation systems by capturing rich interaction context, leading to significant performance gains across various benchmarks.

Generating over 203,000 unique web interaction trajectories, BrowserForge significantly boosts model performance on real-world tasks by leveraging the vastness of the open web.

SeeU redefines 3D rendering by leveraging semantic cues to recover hidden structures, achieving unprecedented quality in sparse-view settings.

PromptPath transforms in-context learning by dynamically tailoring computational pathways, leading to superior performance and interpretability in diverse tasks.

UniPhysGen achieves state-of-the-art performance in grounding physical properties of 3D assets, enabling seamless integration into robotic simulations.

Decoupling perception from reasoning in visual tasks leads to a remarkable 93.2% accuracy on V-Star, showcasing a new paradigm for fine-grained visual reasoning.

LLMs stumble in multi-turn conversations not just because of context length, but because they poison themselves with their own past mistakes – and you can fix it with self-distillation.

InstructSAM equips SAM with high-level instruction understanding and compositional reasoning for multi-instance segmentation, all without modifying SAM's core architecture.

Forget noisy pseudo-labels: SpatialEvo unlocks self-supervised 3D spatial reasoning by generating perfectly accurate training data directly from scene geometry.

Uncertainty-driven zoom-in boosts GUI grounding accuracy by up to 13.4% without any retraining, showing that targeted attention to model uncertainty can significantly improve performance.

Offloading memory and computation to a copilot lets a 7B parameter GUI agent outperform larger models on long-horizon tasks, suggesting a path to more efficient and capable GUI automation.

Finally, a unified open-source framework lets you train, evaluate, and deploy GUI agents across real devices and chat platforms, closing the gap between research and real-world application.

Even frontier models like Claude Sonnet 4.6 stumble when asked to infer user preferences and proactively assist in mobile tasks, achieving less than 50% success despite excelling at explicit task execution.

LLM agents can internalize skills via in-context RL, achieving zero-shot autonomous behavior without the token overhead and retrieval noise of traditional methods.

LLMs can escape the trap of confidently wrong reasoning by co-evolving a generator and verifier from a single model, bootstrapping each other to break free from flawed consensus.

By adversarially co-evolving code and test LLMs, Code-A1 achieves code generation performance on par with human-annotated training, while simultaneously boosting the LLM's ability to find bugs.