Picture for Shiqing Ma

Shiqing Ma

Exposing Product Bias in LLM Investment Recommendation

Add code
Mar 11, 2025
Viaarxiv icon

Holistic Audit Dataset Generation for LLM Unlearning via Knowledge Graph Traversal and Redundancy Removal

Add code
Feb 26, 2025
Viaarxiv icon

Unveiling Provider Bias in Large Language Models for Code Generation

Add code
Jan 14, 2025
Viaarxiv icon

MLLM-as-a-Judge for Image Safety without Human Labeling

Add code
Dec 31, 2024
Figure 1 for MLLM-as-a-Judge for Image Safety without Human Labeling
Figure 2 for MLLM-as-a-Judge for Image Safety without Human Labeling
Figure 3 for MLLM-as-a-Judge for Image Safety without Human Labeling
Figure 4 for MLLM-as-a-Judge for Image Safety without Human Labeling
Viaarxiv icon

Token-Budget-Aware LLM Reasoning

Add code
Dec 24, 2024
Figure 1 for Token-Budget-Aware LLM Reasoning
Figure 2 for Token-Budget-Aware LLM Reasoning
Figure 3 for Token-Budget-Aware LLM Reasoning
Figure 4 for Token-Budget-Aware LLM Reasoning
Viaarxiv icon

Continuous Concepts Removal in Text-to-image Diffusion Models

Add code
Nov 30, 2024
Viaarxiv icon

Bias Similarity Across Large Language Models

Add code
Oct 15, 2024
Viaarxiv icon

Speculative Coreset Selection for Task-Specific Fine-tuning

Add code
Oct 02, 2024
Viaarxiv icon

Data-centric NLP Backdoor Defense from the Lens of Memorization

Add code
Sep 21, 2024
Figure 1 for Data-centric NLP Backdoor Defense from the Lens of Memorization
Figure 2 for Data-centric NLP Backdoor Defense from the Lens of Memorization
Figure 3 for Data-centric NLP Backdoor Defense from the Lens of Memorization
Figure 4 for Data-centric NLP Backdoor Defense from the Lens of Memorization
Viaarxiv icon

Unlocking Adversarial Suffix Optimization Without Affirmative Phrases: Efficient Black-box Jailbreaking via LLM as Optimizer

Add code
Aug 21, 2024
Figure 1 for Unlocking Adversarial Suffix Optimization Without Affirmative Phrases: Efficient Black-box Jailbreaking via LLM as Optimizer
Figure 2 for Unlocking Adversarial Suffix Optimization Without Affirmative Phrases: Efficient Black-box Jailbreaking via LLM as Optimizer
Figure 3 for Unlocking Adversarial Suffix Optimization Without Affirmative Phrases: Efficient Black-box Jailbreaking via LLM as Optimizer
Figure 4 for Unlocking Adversarial Suffix Optimization Without Affirmative Phrases: Efficient Black-box Jailbreaking via LLM as Optimizer
Viaarxiv icon