9 papers · 1 filter
SemTrace: Source-Grounded Semantic Signatures for Tracing LLM Exposure to Protected Documents
Junyan Zhang, Yudong Zeng, Yongwei Huang +3
Large language models are increasingly used to read documents and produce downstream text, creating a provenance problem when the document owner cannot control or inspect the model…
Decoding by Perturbation: Mitigating MLLM Hallucinations via Dynamic Textual Perturbation
Sihang Jia, Shuliang Liu, Songbo Yang +3
Multimodal Large Language Models frequently suffer from inference hallucinations, partially stemming from language priors dominating visual evidence. Existing training-free mitigat…
Correct Prediction, Wrong Steps? Consensus Reasoning Knowledge Graph for Robust Chain-of-Thought Synthesis
Zipeng Ling, Shuliang Liu, Seonil Son +4
Large language models (LLMs) have become increasingly used for various tasks, often coupled with Chain-of-Thought (CoT) prompting to boost accuracy. Recent work has shown that high…
Quantifying LLM Biases Across Instruction Boundary in Mixed Question Forms
Zipeng Ling, Shuliang Liu, Yuehao Tang +8
Large Language Models (LLMs) annotated datasets are widely used nowadays, however, large-scale annotations often show biases in low-quality datasets. For example, Multiple-Choice Q…
LLM Abstention Can Be a Prompt Artifact, in Addition to Genuine Uncertainty
Zipeng Ling, Shuliang Liu, Yuehao Tang +8
Large Language Models (LLMs) are increasingly trained to abstain from answering questions they are unsure about. However, this ability is often misused: in real-world applications,…
Unveiling Instruction-Specific Neurons & Experts: An Analytical Framework for LLM's Instruction-Following Capabilities
Junyan Zhang, Yubo Gao, Yibo Yan +8
The finetuning of Large Language Models (LLMs) has significantly advanced their instruction-following capabilities, yet the underlying computational mechanisms driving these improv…