8 papers
Retrieved But Not Reliable: A Survey on Attacks, and Defenses in Retrieval-Augmented Generation
Minh Tran, Cuong Dang, Tuc Nguyen +10
Retrieval-Augmented Generation (RAG) enhances large language models by grounding outputs in external knowledge, improving factuality and reducing hallucinations. At the same time,…
Spectral Flattening Is All Muon Needs: How Orthogonalization Controls Learning Rate and Convergence
Tien-Phat Nguyen, Truong Nguyen, Minh-Phuc Truong +3
Muon orthogonalizes the momentum buffer before each update, replacing its singular values with ones via Newton-Schulz iterations. This simple change lets Muon tolerate far larger l…
ATLAS: Verifier-Guided Adaptive Latent Activation Steering for Efficient LLM Reasoning
Tuc Nguyen, Thai Le
Recent work on activation and latent steering has demonstrated that modifying internal representations can effectively guide large language models (LLMs) toward improved reasoning…
ShareChat: A Dataset of Chatbot Conversations in the Wild
Yueru Yan, Tuc Nguyen, Bo Su +2
By evaluating Large Language Models (LLMs) through uniform, text-only interfaces, current academic benchmarks obscure how the unique designs and affordances of distinct commercial…
Unraveling Interwoven Roles of Large Language Models in Authorship Privacy: Obfuscation, Mimicking, and Verification
Tuc Nguyen, Yifan Hu, Thai Le
Recent advancements in large language models (LLMs) have been fueled by large scale training corpora drawn from diverse sources such as websites, news articles, and books. These da…
NoMatterXAI: Generating "No Matter What" Alterfactual Examples for Explaining Black-Box Text Classification Models
Tuc Nguyen, James Michels, Hua Shen +1
In Explainable AI (XAI), counterfactual explanations (CEs) are a well-studied method to communicate feature relevance through contrastive reasoning of "what if" to explain AI model…
Generalizability of Mixture of Domain-Specific Adapters from the Lens of Signed Weight Directions and its Application to Effective Model Pruning
Tuc Nguyen, Thai Le
Several parameter-efficient fine-tuning methods based on adapters have been proposed as a streamlined approach to incorporate not only a single specialized knowledge into existing…
Adapters Mixup: Mixing Parameter-Efficient Adapters to Enhance the Adversarial Robustness of Fine-tuned Pre-trained Text Classifiers
Tuc Nguyen, Thai Le
Existing works show that augmenting the training data of pre-trained language models (PLMs) for classification tasks fine-tuned via parameter-efficient fine-tuning methods (PEFT) u…