Yinglin Wang

ReVision: Reprogramming Image Generators for Dense Visual Perception

Recent image generators exhibit rich visual priors that can support dense visual perception. However, existing approaches typically encode multiple perception tasks into a shared parameterization, making joint training sensitive to the mixture and weighting of heterogeneous data and prone to cross-task interference. To...

ValuePEFT: an effective method for steerable multi-level human values in LLMs

V-RoLoRA: RLVR-Driven MoE Routing for Steerable Pluralistic Alignment

This work studies value-controllable alignment through discrete condition vectors and proposes Verifiable-reward-Routed LoRA—a parameter-efficient mixture-of-experts LoRA framework enhanced with conditioned gating, which consistently out-performs prompt-based steering and multi-task PEFT baselines.