MALTO at SemEval-2024 Task 6: Leveraging Synthetic Data for LLM Hallucination Detection
This work addresses the issue of hallucinations in NLG for AI systems, but it appears incremental as it builds on existing NLI methods and data augmentation techniques.
The paper tackled the problem of detecting hallucinations in text generated by large language models by introducing a data augmentation pipeline with LLM-assisted pseudo-labeling and sentence rephrasing, and a voting ensemble of NLI-based models, achieving results in the SHROOM challenge.
In Natural Language Generation (NLG), contemporary Large Language Models (LLMs) face several challenges, such as generating fluent yet inaccurate outputs and reliance on fluency-centric metrics. This often leads to neural networks exhibiting "hallucinations". The SHROOM challenge focuses on automatically identifying these hallucinations in the generated text. To tackle these issues, we introduce two key components, a data augmentation pipeline incorporating LLM-assisted pseudo-labelling and sentence rephrasing, and a voting ensemble from three models pre-trained on Natural Language Inference (NLI) tasks and fine-tuned on diverse datasets.