5 papers
Hallucination Self-Play: Bootstrapping Reinforced Detector via Evolved Generator
Shiping Yang, Shining Liang, Weihao Liu +4
Identifying faithfulness hallucinations in LLM-generated outputs remains challenging due to the scarcity of high-quality annotated data. Recent work relies on advanced LLMs to synt…
Quantifying and Improving the Robustness of Retrieval-Augmented Language Models Against Spurious Features in Grounding Data
Shiping Yang, Jie Wu, Wenbiao Ding +7
Robustness has become a critical attribute for the deployment of RAG systems in real-world applications. Existing research focuses on robustness to explicit noise (e.g., document s…
PIKA: Expert-Level Synthetic Datasets for Post-Training Alignment from Scratch
Shangjian Yin, Shining Liang, Wenbiao Ding +4
High-quality instruction data is critical for LLM alignment, yet existing open-source datasets often lack efficiency, requiring hundreds of thousands of examples to approach propri…
Selected Languages are All You Need for Cross-lingual Truthfulness Transfer
Weihao Liu, Ning Wu, Wenbiao Ding +3
Truthfulness stands out as an essential challenge for Large Language Models (LLMs). Although many works have developed various ways for truthfulness enhancement, they seldom focus…
MuDAF: Long-Context Multi-Document Attention Focusing through Contrastive Learning on Attention Heads
Weihao Liu, Ning Wu, Shiping Yang +4
Large Language Models (LLMs) frequently show distracted attention due to irrelevant information in the input, which severely impairs their long-context capabilities. Inspired by re…