Ninghao Liu

Interpreting and Steering LLMs with Mutual Information-based Explanations on Sparse Autoencoders

GeoBS: Information-Theoretic Quantification of Geographic Bias in AI Models

Investigating CoT Monitorability in Large Reasoning Models

MobileGUI-RL: Advancing Mobile GUI Agent through Reinforcement Learning in Online Environment

MITS: Enhanced Tree Search Reasoning for LLMs via Pointwise Mutual Information

Enhancing LLM Steering through Sparse Autoencoder-Based Vector Refinement

EAP-GP: Mitigating Saturation Effect in Gradient-based Automated Circuit Identification

Concept-Centric Token Interpretation for Vector-Quantized Generative Models

Enhancing Cognition and Explainability of Multimodal Foundation Models with Self-Synthesized Data

Mutual Effort for Efficiency: A Similarity-based Token Pruning for Vision Transformers in Self-Supervised Learning

MQuAKE-Remastered: Multi-Hop Knowledge Editing Can Only Be Advanced with Reliable Evaluations

Interpreting and Steering LLM Representations with Mutual Information-based Explanations on Sparse Autoencoders

CORTEX: Concept-Oriented Token Explanation in Vector-Quantized Generative Model

Receptor-Specific Diffusion Model: Towards Generating Protein-Protein Structures with Customized Perturbing and Sampling