Xinmei Tian

Memory Efficient Fine-Tuning of LLMs via Forward-Only Hessian-Free Coordinate Descent

Detecting Generated Images via Machine Unlearning

MedReason-Dx: Benchmarking Step-by-Step Reasoning of Language Models in Medical Diagnosis

Combining-on-Graph: Structured Reasoning of LLMs with Knowledge Graph

Detecting Generated Images by Fitting Natural Image Distributions

Towards Generalizable Detector for Generated Image

Epistemic Uncertainty for Generated Image Detection

A Theoretical Perspective: How to Prevent Model Collapse in Self-consuming Training Loops

Leveraging Submodule Linearity Enhances Task Arithmetic Performance in LLMs

An Effective Levelling Paradigm for Unlabeled Scenarios

Intention Model: A Novel Explanation for In-context Learning

Overfitting: An Unexpected Asset in AI‐Generated Image Detection

Enhancing Target-unspecific Tasks through a Features Matrix

Prompt Reverse Learning: Enhancing Visual Language Models for Rare Image Recognition

Detecting Discrepancies Between Generated and Natural Images Using Uncertainty

Memory Efficient Block Coordinate Descent Method for Forward-Only Second-Order Finetuning of LLM Models

Consistency Verification for Detecting AI-Generated Images

Memory-Efficient Block Coordinate Descent for Hessian-Informed Zeroth-Order Optimizer

Unveiling the Flow of Input-label Mappings for In-context Learning