Yuhang Liu

A single interaction is all it takes for an adversary to manipulate LLM memory systems, steering outputs toward a pre-specified target with alarming precision.

Reasoning validity in LLMs can be more accurately assessed through a novel three-stream detector that integrates motion with contextual state information, achieving up to 21% better accuracy than existing methods.

Integrating clinical risk factors into deep learning models can dramatically enhance the accuracy of coronary artery stenosis grading, outperforming traditional methods.

Stellar slashes memory overhead and query latency by orders of magnitude, transforming multimodal document retrieval efficiency.

Boundary-region entanglement is a critical bottleneck for GNNs, and our adaptive approach boosts classification accuracy by over 3% while maintaining model stability.

Unseen single-cell perturbation effects can be predicted more accurately by explicitly modeling the latent, dynamic causal processes that drive cellular response.

Open-source Luwen shows that adapting general-purpose LLMs with legal-specific data and retrieval augmentation can yield significant performance gains on complex Chinese legal tasks.

LLMs ace the setup but fumble the execution in mathematical modeling, revealing a critical gap that scaling alone won't fix.

Instruction-following in large reasoning models gets a serious upgrade with RAIN-Merging, a gradient-free technique that merges in instruction-tuned capabilities without wrecking the model's ability to think step-by-step.