Huawei Shen

Fine-tuning Done Right in Model Editing

DIVA-GRPO: Enhancing Multimodal Reasoning through Difficulty-Adaptive Variant Advantage

Latent Reasoning in LLMs as a Vocabulary-Space Superposition

Learning from Mistakes: Negative Reasoning Samples Enhance Out-of-Domain Generalization

GeoVLMath: Enhancing Geometry Reasoning in Vision-Language Models via Cross-Modal Reward for Auxiliary Line Creation

GLOSE: Understanding and Mitigating Toxicity in LLMs via Global Toxic Subspace

Inference-time Alignment in Continuous Space

Cross-Modal Safety Mechanism Transfer in Large Vision-Language Models

A Theory for Token-Level Harmonization in Retrieval-Augmented Generation

Everything is Editable: Extend Knowledge Editing to Unstructured Data in Large Language Models