Jinhe Bi

MINED: Probing and Updating with Multimodal Time-Sensitive Knowledge for Large Multimodal Models

Beyond Magic Words: Sharpness-Aware Prompt Evolving for Robust Large Language Models with TARE

Reinforcement Mid-Training

Reasoning Self-Evaluation via Trajectory Dynamics Modeling

Can visual input Be Compressed? A visual input token compression benchmark for large multimodal models

KORE: Enhancing Knowledge Injection for Large Multimodal Models via Knowledge-Oriented Augmentations and Constraints

HYPERION: Fine-Grained Hypersphere Alignment for Robust Federated Graph Learning

Backdoor Cleaning without External Guidance in MLLM Fine-tuning

Visual Instruction Tuning with 500x Fewer Parameters through Modality Linear Representation-Steering

FedBiP: Heterogeneous One-Shot Federated Learning with Personalized Latent Diffusion Models