Edit model card

Untruthful-Llama2-7B-Bio

We induce hallucinations from the original Llama2-7B by finetuning it on 3.5k hallucinated biographies. We then use it in our ICD method for improve factuality of LLMs and evaluate the effectiveness on TruthfulQA. More details are provided in our Github and Paper.

Downloads last month
4

Dataset used to train HillZhang/untruthful_llama2_7b_bio