Picture for Xuefeng Bai

Xuefeng Bai

From Bias to Balance: Exploring and Mitigating Spatial Bias in LVLMs

Add code
Sep 26, 2025
Viaarxiv icon

Evaluating and Steering Modality Preferences in Multimodal Large Language Model

Add code
May 27, 2025
Viaarxiv icon

Handling Imbalanced Pseudolabels for Vision-Language Models with Concept Alignment and Confusion-Aware Calibrated Margin

Add code
May 04, 2025
Viaarxiv icon

MoK-RAG: Mixture of Knowledge Paths Enhanced Retrieval-Augmented Generation for Embodied AI Environments

Add code
Mar 18, 2025
Viaarxiv icon

Representation-based Reward Modeling for Efficient Safety Alignment of Large Language Model

Add code
Mar 13, 2025
Figure 1 for Representation-based Reward Modeling for Efficient Safety Alignment of Large Language Model
Figure 2 for Representation-based Reward Modeling for Efficient Safety Alignment of Large Language Model
Figure 3 for Representation-based Reward Modeling for Efficient Safety Alignment of Large Language Model
Figure 4 for Representation-based Reward Modeling for Efficient Safety Alignment of Large Language Model
Viaarxiv icon

Adaptive Inner Speech-Text Alignment for LLM-based Speech Translation

Add code
Mar 13, 2025
Viaarxiv icon

XIFBench: Evaluating Large Language Models on Multilingual Instruction Following

Add code
Mar 10, 2025
Viaarxiv icon

Generator-Assistant Stepwise Rollback Framework for Large Language Model Agent

Add code
Mar 04, 2025
Figure 1 for Generator-Assistant Stepwise Rollback Framework for Large Language Model Agent
Figure 2 for Generator-Assistant Stepwise Rollback Framework for Large Language Model Agent
Figure 3 for Generator-Assistant Stepwise Rollback Framework for Large Language Model Agent
Figure 4 for Generator-Assistant Stepwise Rollback Framework for Large Language Model Agent
Viaarxiv icon

The Rise of Darkness: Safety-Utility Trade-Offs in Role-Playing Dialogue Agents

Add code
Feb 28, 2025
Viaarxiv icon

The Power of Personality: A Human Simulation Perspective to Investigate Large Language Model Agents

Add code
Feb 28, 2025
Figure 1 for The Power of Personality: A Human Simulation Perspective to Investigate Large Language Model Agents
Figure 2 for The Power of Personality: A Human Simulation Perspective to Investigate Large Language Model Agents
Figure 3 for The Power of Personality: A Human Simulation Perspective to Investigate Large Language Model Agents
Figure 4 for The Power of Personality: A Human Simulation Perspective to Investigate Large Language Model Agents
Viaarxiv icon